The Development of the Social Self

The Development of the

Social Self

Edited by
Mark Bennett and Fabio Sani
First published 2004
by Psychology Press
27 Church Road, Hove, East Sussex BN3 2FA
Simultaneously published in the USA and Canada
by Psychology Press
29 West 35th Street, New York, NY 10001
This edition published in the Taylor & Francis e-Library, 2004.
Psychology Press is a member of the Taylor & Francis Group
Copyright © 2004 Psychology Press

All rights reserved. No part of this book may be reprinted or

reproduced or utilised in any form or by any electronic, mechanical, or
other means, now known or hereafter invented, including photocopying
and recording, or in any information storage or retrieval system,
without permission in writing from the publishers.
British Library Cataloguing in Publication Data
A catalogue record for this book is available
from the British Library
Library of Congress Cataloging-in-Publication Data

ISBN 0-203-39109-8 Master e-book ISBN

ISBN 0-203-67781-1 (Adobe eReader Format)

ISBN 1-841-69294-8 (Print Edition)
 TO OUR BELOVED SONS,
NICHOLAS AND LEONARDO
Contents

List of tables xi
List of figures xii
List of contributors xiii

1 Introduction: Children and social identity 1

MARK BENNETT AND FABIO SANI

Social and developmental perspectives on the self-concept 2

On the benefits of a closer association between social and
developmental psychology 4
The social identity approach 6
Organisation of the book 17
References 22

PART I
Basic issues

2 The development of a sense of “we”: The emergence and

implications of children’s collective identity 29
DIANE N. RUBLE, JEANNETTE ALVAREZ, MEREDITH BACHMAN,
JESSICA CAMERON, ANDREW FULIGNI, CYNTHIA GARCIA COLL,
AND EUN RHEE

The nature of social identity 30

When and how do children understand and identify with social
categories? 33
Developmental changes in the significance of collective
identity 42
Consequences of developmental changes in collective
identity 51
Summary and conclusions 63
References 64
 Contents vii
3 Developmental aspects of social identity 77
FABIO SANI AND MARK BENNETT

Children’s conception of social identities 78

The flexibility of children’s understanding of in-group
identity 85
The emergence of social identity 91
General summary and conclusions 95
References 97

PART II
Identities

4 Gender as a social category: Intergroup processes and

gender-role development 103
KIMBERLY K. POWLISHTA

Intergroup processes 104

The development of gender categories 107
Exaggeration of between-sex differences and within-sex
similarities 108
Own-sex favouritism 112
The impact of social context on the salience of gender 113
Individual and group differences in gender-based intergroup
processes 115
How is gender different from other social categories? 122
Conclusions and implications for improving cross-sex
relations 125
References 126

5 The gender wars: A self-categorisation theory perspective on

the development of gender identity 135
BARBARA DAVID, DIANE GRACE, AND MICHELLE K. RYAN

The developmental sequence 136

Theories of gender development 139
Matching the theories and the evidence 140
The self as a process 141
Self-categorisation, the dominant theories, and gender
development 145
Summary and suggestions 150
References 152
viii Contents
6 The development of national identity and social identity
processes: Do social identity theory and self-categorisation
theory provide useful heuristic frameworks for developmental
research? 159
MARTYN BARRETT, EVANTHIA LYONS, AND ARANTZA
DEL VALLE

The principal phenomena that characterise the development of

national identity 160
The relevance of SIT and SCT to the study of national identity
development 161
The application of SIT and SCT to the development of national
identity 164
Evaluation of the predictions 165
So do SIT and SCT provide useful heuristic frameworks for
developmental research? 182
References 185

7 Ethnic identity and social context 189

MAYKEL VERKUYTEN

From social cognition to situation 190

Intergroup theories 192
Perceived discrimination 195
Cultural orientations 199
Ethnic self-esteem and context 201
Ethnic self-categorisation 202
Ethnic self-categorisation and behaviour 203
Conclusion 207
References 210

PART III
Applications

8 Social identity processes and children’s ethnic prejudice 219

DREW NESDALE

Explanations of children’s prejudice 220

Social identity theory and ethnic prejudice 225
Social identity development theory of children’s ethnic
prejudice 226
Conclusions 238
References 240
 Contents ix
9 The development and self-regulation of intergroup attitudes
in children 247
ADAM RUTLAND

Children’s intergroup attitudes and cognitive-developmental

theory 247
Recent research on children’s national intergroup
attitudes 249
Social identity theory 251
Self-regulation and children’s intergroup attitudes 252
Implicit measures of children’s intergroup attitudes 254
Dissociation between implicit and explicit measures 255
The Implicit Association Test 257
Public self-focus and children’s ethnic intergroup
attitudes 258
Conclusion 260
References 261

10 Reducing stepfamily conflict: The importance of inclusive

social identity 267
BRENDA S. BANKER, SAMUEL L. GAERTNER, JOHN F. DOVIDIO,
MISSY HOULETTE, KELLY M. JOHNSON, AND BLAKE M. RIEK

The role of social categorisation and intergroup

relations 269
Reducing intergroup bias and conflict 271
Stepfamily marriages 276
Conclusions 283
References 285

PART IV
Epilogue

11 The development of social identity: What develops? 291

DOMINIC ABRAMS

The chapters 291

Conclusions 305
References 307
x Contents
12 Towards a developmental social psychology of the social self 313
KEVIN DURKIN

The family as the first group 314

Developmental changes and continuities 315
Competition between developmental and social processes 317
Language 319
Age as a social identity 320
Social identities are diverse – and may develop differently 321
Can we create better social selves? 322
Social selves through the lifespan 323
Conclusions 324
References 325

Author index 327

Subject index 335

Tables

6.1 Evaluations of the national in-group versus various national

out-groups 167
6.2 Affect expressed towards the national in-group versus
various national out-groups 169
6.3 Evaluation of each individual national group and either the
importance of national identity or degree of national
identification 171
6.4 Positive distinctiveness of the national in-group and either
the importance on national identity or degree of national
identification 172
6.5 Affect expressed towards each individual national group
and either the importance of national identity or degree of
national identification 173
6.6 Affective distinctiveness of the national in-group and either
the importance of national identity or degree of national
identification 174
6.7 Perceived variability of each individual national group and
either the importance of national identity or degree of
national identification 179
6.8 Importance of national identity, degree of national
identification, and perceived variability of the national
in-group 180
Figures

10.1 Longitudinal analysis: Contact and conflict 281

10.2 Longitudinal analysis: Contact and one family 282
10.3 Longitudinal analysis: One family and conflict 282
Contributors

Dominic Abrams, Department of Psychology, University of Kent, Canterbury,

Kent CT2 7NP, UK
Jeannette Alvarez, Department of Psychology, New York University,
6 Washington Place, 419, New York, NY 10003, USA
Meredith Bachman, Department of Psychology, New York University,
6 Washington Place, 7th Floor, New York, NY 10003, USA
Brenda S. Banker, Department of Psychology, University of Delaware,
Newark, DE 19716, USA
Martyn Barrett, Department of Psychology, University of Surrey, Guildford
GU2 7XH, UK
Mark Bennett, Department of Psychology, University of Dundee, Dundee
DD1 4HN, Scotland, UK
Jessica Cameron, Department of Psychology, New York University,
6 Washington Place, 4th Floor, New York, NY 10003, USA
Barbara David, School of Psychology, The Australian National University,
Canberra ACT 0200 Australia
Arantza del Valle, Departament de Psicologia, Universitat de Girona, Placa
Sant Domenech, 9, 17071 Girona, Spain
John F. Dovidio, Department of Psychology, Colgate University, Hamilton,
NY 13346, USA
Kevin Durkin, School of Psychology, University of Western Australia, Craw-
ley, Western Australia 6009
Andrew J. Fuligni, Center for Culture and Health, UCLA Neuropsychiatric
Institute, 760 Westwood Plaza, Box 62, Los Angeles, CA 90024-1759,
USA
Cynthia Garcia Coll, Education Department, Brown University, P.O. Box
1938, Providence, RI 02912, USA
xiv Contributors
Samuel L. Gaertner, Department of Psychology, University of Delaware,
Newark, DE 19716, USA
Diane Grace, School of Psychology, Australian National University,
Canberra ACT 0200, Australia
Missy Houlette, Department of Psychology, University of Delaware,
Newark, DE 19716, USA
Kelly M. Johnson, Department of Psychology, University of Delaware,
Newark, DE 19716, USA
Evanthia Lyons, Department of Psychology, University of Surrey, Guildford
GU2 7XH, UK
Drew Nesdale, School of Applied Psychology, Gold Coast Campus, Griffith
University, PMB 50, Queensland 9726, Australia
Kimberly K. Powlishta, Department of Psychology, Saint Louis University,
3511 Laclede Avenue, St. Louis, MO 63103-2010, USA
Eun Rhee, Department of Psychology, University of Delaware, Newark, DE
19716, USA
Blake M. Riek, Department of Psychology, University of Delaware, Newark,
DE 19716, USA
Diane N. Ruble, Department of Psychology, New York University, 6
Washington Place, 751, New York, NY 10003, USA
Adam Rutland, Department of Psychology, University of Kent at Canterbury,
Canterbury, Kent CT2 7NP, UK
Michelle K. Ryan, School of Psychology, Australian National University,
Canberra ACT 0200, Australia
Fabio Sani, Department of Psychology, University of Dundee, Dundee DD1
4HN, Scotland, UK
Maykel Verkuyten, Department of Social Sciences, Utrecht University, PO
Box 80140, 3508 TC Utrecht, The Netherlands
1 Introduction: Children and
social identity
Mark Bennett and Fabio Sani

The developing self has been a source of intermittent interest to psycholo-

gists ever since the early pioneering contributions of Baldwin (1895), Cooley
(1902), James (1890), and Mead (1934). Though much research has been
directed at the development of that aspect of the self-concept referred to as
the personal self (the self defined by idiosyncratic features, such as personality
traits), the aim of this book is to consider the development of the social self,
that is, the self defined by one’s membership of social groups – for example,
gender, age, ethnicity, nationality, religion, and subcultural groups. Broadly,
our project is undertaken within the framework afforded by the social
identity approach (Tajfel, 1972, 1978; Tajfel & Turner, 1979; Turner, 1975,
1982), a social-cognitive perspective that sees the self-concept as the outcome
of a self-categorisation process. At its heart is an acknowledgement of the
interdependence of self and social context. The chapters of this volume
address the development of a range of phenomena that fall within the pur-
view of the social identity perspective – for example, social categorisation,
self-conception, social comparison, and prejudice.
One of the central insights of the social identity approach is that groups
can become part of the self-concept. This contention contrasts starkly
with developmental psychologists’ almost exclusive treatment of groups as
external to the self, as merely an influence upon the self. Within the social
identity approach, the self is taken to comprise both personal and social
identity, and neither is seen as in any sense more fundamental or authentic
than the other. Theoretically, treating groups as constitutive of the self is
important in that it becomes possible to specify relations between self-
conceptions and many aspects of social behaviour; moreover, it is to recog-
nise that our social identity is associated with actions and cognitions that are
discontinuous with those that arise from personal identity.
In this chapter we note the case for a developmental approach to the study
of social identity. Briefly, we also consider the complementarity of social
psychological and developmental approaches to the self. Following this we
provide an introduction to the theoretical position underpinning this volume,
that is, the social identity approach. Finally, we outline the content of the
book’s chapters.
2 Introduction
SOCIAL AND DEVELOPMENTAL PERSPECTIVES ON THE
SELF-CONCEPT

A major impetus for this volume has been the recognition, on the one hand,
that social psychologists have neglected the development of self, and on the
other, that developmentalists have focused largely upon the personal self. That
is, most research on the development of the self-concept has been conducted
by developmentalists and has addressed children’s conceptions of themselves
in terms of idiosyncratic attributes, such as personality traits, abilities,
and preferences. (See, for example, Harter, 1998, 1999, for comprehensive
reviews.) Little attention has been directed at the development of the social
self. Nonetheless, it is clear that children are de facto members of many social
groups; they are also members of groups that they themselves create. In our
view, it is important to understand when and how social categories become
constitutive of the self, such that children subjectively identify with them. As
Ruble et al. (Chapter 2) put it, we need to understand children’s developing
sense of “we”. In addition to its importance at the level of description,
the developmental study of the social self is important insofar as it is clear
that, among adults, social identities mobilise specific forms of group-related
action and perception, such as cooperation with and attraction to in-group
members (e.g., see Hogg & Abrams, 1988). As we see it, a key aim in the
longer term is to understand the ontogenesis of the relation between social
self-conceptions and social action. More generally, the focus upon children’s
developing social identities seems a significant enterprise given that one’s
social identity “creates and defines the individual’s place in society” (Tajfel &
Turner, 1979, pp. 40–41).
To address the development of the social self, we believe, necessitates
reference to both social and developmental psychology. However, a backward
glance at self research in developmental and social psychology reveals a
striking degree of disciplinary insularity. A cursory inspection of papers in
the two disciplines shows that, apart from standard genuflections towards
founding figures, such as Baldwin, Cooley, James, and Mead, references are
largely nonoverlapping. This is perhaps surprising given clear parallels in
terms of the sorts of topics that have been studied. For example, both
developmental and social psychologists have examined self-esteem, self-
efficacy, and, most conspicuously of all, self-representations. And both have
adopted broadly cognitive views of the self, reflecting a commitment, whether
implicit or explicit, to Mead’s (1934) fundamental insight that “Self-
consciousness, rather than affective experience [. . .], provides the core and
primary structure of the self, which is thus essentially a cognitive rather than
an emotional phenomenon” (p. 173).
Despite obvious disciplinary similarities at the level of substantive inter-
ests, there have nonetheless been considerable differences in terms of theor-
etical orientation and methods. For example, considering the two fields’ study
of self-representations, quite contrasting traditions are apparent. Within
 Bennett and Sani 3
social psychology, cognitive approaches to the self have been highly influen-
tial at least since Markus’ (1977) introduction of the concept of self-schemas,
that is, knowledge structures pertaining to the self. (See also Higgins, 1987;
Kihlstrom & Cantor, 1984.) Among developmentalists, too, the self has
been conceived in cognitive terms and a considerable amount of attention
has been given to self-representations (Harter, 1999). However, the develop-
mental study of the self has been embedded in an entirely different tradition:
that of cognitive-developmental theory – specifically, Piagetian and neo-
Piagetian theory (e.g., Case, 1992; Fischer, 1980). Here, a guiding assumption
has been that general conceptual development plays a central role in
determining the emergence, form, and increasing differentiation of self-
conceptions. That is, researchers have sought to account for age-related
structural growth in self-conceptions in terms of cognitive ontogenesis.
A further difference is that social psychologists, unlike most developmen-
talists, have asked questions about process-related issues. For example, what
are the motivations underlying particular social identities (e.g., Brewer, 1999;
Deaux, Reid, Mizrahi, & Cotting, 1999), and how do social identities guide
particular types of action (e.g., Reicher, 1984)? Developmental psychologists,
however, have more typically sought to describe the changing contents of
children’s self-conceptions (e.g., Damon & Hart, 1988; Keller, Ford &
Meacham, 1978).
A further and key difference is that within social psychology, particularly
under the influence of social identity theory, social self-conceptions (i.e.,
social identities) have been studied insofar as they are hypothesised to be
consequential with respect to a broad range of inter- and intragroup phe-
nomena. That is, social identity theorists have examined the relationship
between self-conception and many forms of social behaviour, such as
cooperation, conformity, crowd behaviour, group polarisation, and so on.
Developmentally oriented work, however, has typically sought to describe
developing self-conceptions as an end in themselves, or as outcomes of par-
ticular socialisation experiences (although for exceptions, see for example
Abrams, Rutland, Cameron, & Marques, 2003; Bigler, Jones, & Lobliner,
1997; Harter, 1998, 1999).
A more general difference between the developmental and social psycho-
logical study of the self is in the sheer range of topics studied: Unsurprisingly,
perhaps, given the greater prominence of the self-concept in social psy-
chology, many more topics have been studied here than within developmental
psychology. Thus, broadly speaking, developmental psychologists have been
concerned primarily (though not exclusively) with the initial emergence of the
self-concept (Brooks-Gunn & Lewis, 1979; Rochat, 2001), the development
of, and social influences upon, self-representations (Harter, 1999), and self-
esteem (Harter, 1987). Social psychologists, however, have addressed a pro-
digious range of self-related topics, including self-affirmation (Steele, 1988),
self-awareness (Fenigstein, Scheier, & Buss, 1975), self-complexity (Linville,
1985), self-discrepancy (Higgins, 1987), self-enhancement (Wills, 1981),
4 Introduction
self-handicapping (Jones & Berglas, 1978), self-evaluation maintenance
(Tesser, 1988), self-monitoring (Snyder, 1974) self-representations (Markus,
1977) self-presentation (Schlenker, 1980), self-verification (Swann & Read,
1981), and so on. Nonetheless, there are signs that the developmental study
of the self is poised to become an increasingly significant focus of research
(as reflected, for example, in major symposia at international conferences and
recent research-based volumes, e.g., Brandstädter & Lerner, 1999; Demetriou
& Kazi, 2001; Harter, 1999; Moore & Lemmon, 2001).

ON THE BENEFITS OF A CLOSER ASSOCIATION

BETWEEN SOCIAL AND DEVELOPMENTAL PSYCHOLOGY

Such differences as we have thus far identified to some extent reflect differ-
ences in the disciplines’ orientations. Social psychologists are interested in
processes of social influence, conceived in its broadest sense (see Allport,
1968; McGarty & Haslam, 1997). Developmentalists, however, focus primar-
ily on change over time. These differences in orientation suggest possible
benefits of closer association between the two disciplines, something noted by
many before us (e.g., Brehm, Kassin, & Gibbons, 1981; Durkin, 1995; Eckes
& Trautner, 2000; Flavell & Ross, 1981; Masters & Yarkin-Levin, 1984). Ini-
tially, let us turn to possible contributions of developmental psychology to
social psychology.
Typically, social psychology has not dealt well with the issue of change,
either generally or in the specific field of the study of self. Nor has it been
much interested in the particular contents of self-representations. However, it
seems reasonable to suggest that important changes in the content of the
self-concept occur over the life-course, and that these may have significant
implications for many aspects of psychological functioning. As Eckes and
Trautner (2000) have commented, “transitions such as reaching puberty,
becoming a parent, or retiring can be conceptualised as sensitive periods
systematically influencing an individual’s self-construal” (p. 7). Thus, an
appreciation of the particular contents of self-representations and how they
change is likely to be central to an understanding of self-functioning during
the various phases of adulthood. Clearly, developmental psychology offers
much in terms of the analysis and measurement of change (e.g., Brandstädter
& Lerner, 1999; Lerner, 1998) and may represent a valuable resource to self
theorists working within social psychology.
In addition to highlighting the need to look at development during adult-
hood, there is a need too to understand early origins of self-processes. “It is
elementary yet widely overlooked by social psychologists that the phenomena
they study do not arise out of nowhere, forming miraculously just before their
subjects come to university” (Durkin, 1995, p. 3). That is, we need to acknow-
ledge the developmental histories associated with social processes, specifying
origins and developmental trajectories. Moreover, the possible antecedents
 Bennett and Sani 5
of aspects of adult functioning in childhood experiences could profitably be
addressed. In the absence of an appreciation of these sorts of distal causes, as
Durkin comments, social psychology “risks becoming a science of proximal
effects” (p. 6).
Recently, Pomerantz and Newman (2000) have argued forcefully that
“attention to developmental psychology would enrich research programs
within social psychology by providing new perspectives on the issues with
which the field grapples” (p. 301). For example, they suggest that the study
of developmental origins can provide insights into individual differences.
They also discuss how young children’s initial reactions to classes of social
stimuli may provide a window onto what become basic and automatic
responses in adulthood. At a methodological level they note that replica-
tion on samples of children (invariably more representative than samples
based on undergraduate populations; Sears, 1986) provides evidence for the
robustness of effects. Thus, developmental research can provide a powerful
source of theory confirmation. And in the absence of confirmatory evi-
dence (i.e., where children’s behaviour is found to differ from that of
adults), we must question the assumption of a theory’s universality; the
challenge to understand developmental origins and change then becomes
pressing.
Turning to the benefit to developmental psychology of a closer acquaint-
ance with social psychology, we suggest that a fundamental gain will be a
better understanding of cognitive and behavioural variability over social
contexts – a theme that emerges in many of the chapters of this volume.
A key feature of much social psychological theorising (and particularly
self-categorisation theory) is the attempt to specify relationships between
contextual variables and intra-individual variability in social behaviour and
cognition, including, of course, self-related phenomena. A central goal,
then, is to account for the fact that individuals’ behaviour and cognition
covaries with social contexts. Generally speaking, at the intra-individual
level, developmental psychologists have not accorded context an essential
role1: In looking primarily at change over time, change over contexts has
frequently been overlooked. Where contextual variation has been examined,
this has often been in an ad hoc empirical way, rather than in a more prin-
cipled and theorised way, as is typical in social psychology generally, and
within the social identity approach particularly.
In seeking to understand the development of the self, it is apparent that
social and developmental psychology each provide important insights. Social
psychologists, while neglecting processes of change over time, give attention
to social processes within given contexts; developmentalists, however, focus
upon change over time, often disregarding changes over contexts. Thus, an

1 This is patently not true at the inter-individual level, where context has been accorded a
foundational role in development, especially from a Vygotskian perspective.
6 Introduction
approach that integrates the strengths of the two disciplines may be fertile
indeed. However, as Eckes and Trautner (2000) have argued, an integrative
approach “must not confine itself to simply adding to the first perspective
what the second has to offer and vice versa. Quite the contrary. At the
intersection of developmental and social psychology many issues will emerge
that pose new kinds of challenges for theorizing and research” (p. 12). In
particular, a social-developmental perspective should aspire to explore the
possible ways in which context varies with age, since the impact of contexts
upon self-processes is likely to be importantly mediated by cognitive-
developmental factors. Following Eckes and Trautner, then, we suggest that
the social-developmental study of the self must recognise that the self is
subject to both social and developmental processes, and that these processes
are likely to be reciprocally influential.
Our discussion thus far implies an ambitious programme of work on the
development of the social self. This book represents an attempt to make a
start on this project, bringing together some of the best current work in this
new area of inquiry. In seeking to explore children’s social selves, we take the
view that the social identity approach represents a valuable theoretical
resource. It is a theoretical perspective that has had a colossal impact on the
social-psychological literature; if there is a more promising theoretical basis
for the investigation of the developing social self, we are unaware of it. For
our developmental readership, to whom the social identity approach may be
relatively unfamiliar, we now provide a brief statement of the origins and
major features of the two main theories that comprise the social identity
approach.

THE SOCIAL IDENTITY APPROACH

During the late 1960s and early 1970s, both the theoretical and the method-
ological foundations of social psychology were shaken by a wave of criticism
coming from social psychologists themselves (Gergen, 1973; McGuire, 1973;
Ring, 1967). A pervasive sense of dissatisfaction rapidly spread throughout
the community, causing what has been characterised as a “crisis of con-
fidence” (Elms, 1975). It was in this climate that a fast-growing group of
European social psychologists – led by Henri Tajfel in Britain and Serge
Moscovici in France – attempted to create a “new look” social psychology.
The declared aim of this group of scholars was to revalue and emphasise the
“social” dimension of human behaviour (Doise, 1982; Jaspars, 1980; Tajfel,
1984) by creating an anti-individualistic and anti-reductionistic social
psychology.
As far as the specific field of group processes was concerned, the anti-
reductionistic stance was represented by what we now know as the “social
identity approach” (Hogg & Abrams, 1988), which developed throughout the
1970s and the 1980s. This approach now comprises two main theories, social
identity theory and latterly self-categorisation theory.
 Bennett and Sani 7
Social identity theory
The initial focus of the social identity approach was the study of intergroup
behaviour (Billig, 1976; Tajfel, 1970; Tajfel, Flament, Billig, & Bundy, 1971),
which resulted in the creation of a formally stated theory that became known
as social identity theory, or SIT for short (Tajfel & Turner, 1986). The explicit
objective of SIT was that of generating a truly social-psychological perspec-
tive on intergroup relations, as opposed to the dominant individualistic out-
look. The metatheoretical distinction between individualistic theories and
social psychological theories of group processes and intergroup relations is
clearly summarised in the following statement:

Many of the ‘individual’ theories start from general descriptions of psy-

chological processes which are assumed to operate in individuals in a way
which is independent of the effects of social interaction and social con-
text. The social context and interaction are assumed to affect these pro-
cesses, but only in the sense that society provides a variety of settings in
which the ‘basic’ individual laws of motivation or cognition are uni-
formly displayed. In contrast, ‘social psychological’ theories [. . .] stress
the need to take into account the fact that group behaviour – and even
more so intergroup behaviour – is displayed in situations in which we are
not dealing with random collections of individuals who somehow come
to act in unison because they all happen to be in a similar psychological
state.
(Tajfel, 1981, p. 403)

In order to appreciate the empirical base from which the theory emerged, it is
necessary to discuss some of the experiments conducted by Tajfel and his
colleagues during the early 1970s.

The “minimal groups experiments”

The impetus for the development of SIT was the need to explain unexpected
results emerging from some experiments conducted by Tajfel and his
colleagues at Bristol University in the early 1970s. Tajfel was interested in
intergroup discrimination, and was intrigued by the fact that various studies
had cast doubts on Sherif’s (1967) assumption that realistic intergroup com-
petition is the necessary condition for conflict and ethnocentrism. For
example, Ferguson and Kelley (1964) had found that even in the absence of
competition, people tend to evaluate the performance of the in-group more
favourably than that of the out-group. Other authors (Rabbie & Horwitz,
1969; Rabbie & Wilkens, 1971) had established that the simple anticipation
of future interaction between groups may determine negative bias in the
participants’ evaluation of the out-group.
In the light of such findings, Tajfel decided to investigate the minimal
8 Introduction
necessary condition for intergroup discrimination. In order to do so, he set up
a series of experiments reproducing a situation in which both the character-
istics of the two groups and the nature of their relation were minimal, in
the sense that they lacked all the common characteristics of real contexts.
That situation was supposed to constitute a “no discrimination” baseline to
which other variables could be cumulatively added in order to investigate the
necessary preconditions for the emerging of discriminative and ethnocentric
behaviour. One of these experiments turned out to be crucial, and was
successfully replicated several times and by different authors.
In this experiment, 14- to 16-year-old boys were assigned to one of two
groups, ostensibly on the basis of their preferences for paintings by either
Klee or Kandinsky. In fact, assignment was random. Subsequently, each par-
ticipant was placed in a separate cubicle, and was told to which group – either
“Klee” or “Kandinsky” group – he belonged. Participants had no contact
with other in-group members, nor with the out-group. They were then
requested to allocate points representing money to various participants in the
experiment, using specially prepared decision matrices. Each matrix allowed
allocation of money to two recipients. The recipients were anonymous, but
their group affiliation was revealed: one was an in-group member and the
other an out-group member. However, the participants were told that
giving money to the in-group members would not contribute to what they
themselves would receive.
The structure of the matrices was such that it was possible to assess the
strategies adopted by the participants for the allocation of points. The strat-
egies were: (a) maximising the total number of points for the two recipients
irrespective of their group membership; (b) maximising the number of points
for in-group members; (c) maximising the difference in favour of the in-group
members in the number of points allocated; (d) distributing an equal number
of points to the members of the in-group and those of the out-group.
Because of the “minimal” nature of the intergroup context, Tajfel and his
colleagues were expecting not to find any discrimination between groups;
indeed, this situation had been conceived as a baseline against which the impact
of further variables could be judged. However, the results came as a surprise in
that the participants showed a clear tendency to discriminate against the out-
group by assigning more money to in-group members. Most interestingly,
their concern was not with maximising the absolute profit of the in-group,
but with maximising the difference between the profit of the in-group and that
of the out-group in favour of the in-group. Thus, rather than adopting a simple
economic strategy (“Let’s get as much as we can”), they adopted a strategy
that ensured superiority over the out-group (“Let’s make sure that we do
better than them”). In other words, mere categorisation in terms of in-group/
out-group proved to be connected with the emergence of discrimination.
These results were initially explained in terms of “norms” prevalent in
some societies (Tajfel et al., 1971). However, Tajfel and colleagues soon
rejected this explanation: It was seen as uninteresting in the sense that it was
 Bennett and Sani 9
not genuinely heuristic, since whatever the result of the experiment, an
explanation in terms of norms could be invoked. Therefore the central
question was still unanswered: Why does discrimination follow from categor-
isation? Is discrimination a straightforward consequence of the process of
social categorisation, or is there a more articulated story to be revealed? SIT
sought to answer this question.

The core features of social identity theory

The theory elaborated by Tajfel is underpinned by three basic and intercon-
nected concepts: social identity, social categorisation, and social comparison.
Social identity – which self-categorisation theory eventually distinguished
from personal identity (Turner & Reynolds, 2001) – is defined as “that part of
an individual’s self-concept which derives from his knowledge of his member-
ship of a social group (or groups) together with the value and emotional
significance attached to that membership” (Tajfel, 1981, p. 255). The identifi-
cation of the self with a social group, and therefore the acquisition of a social
identity, goes hand-in-hand with the process of social categorisation.
Social categorisation refers to the cognitive segmentation of the social
environment into different social categories. This operation both systematises
the social world and provides a system of orientation for the self, by creating
and defining the individual’s place in society. In other words, an individual in
a given social situation could not identify him/herself with a certain social
group without psychologically structuring the context in terms of relatively
discrete social categories. The tendency to segment the social environment
into distinct social units is partly made possible by the peculiar characteristics
of the process of categorisation. In turn, categorisation of stimuli – either
natural or social – leads to the cognitive accentuation both of differences
between stimuli belonging to different categories and of similarities between
stimuli belonging to the same category (Doise, 1978; Eiser & Stroebe, 1972;
Tajfel & Wilkes, 1963).
Social comparison refers to the tendency to evaluate the categories consti-
tuting the context by comparing them on relevant dimensions. This process
contributes fundamentally to an understanding of the social world and the
in-group within it. As Tajfel has argued, “the characteristics of one’s group
as a whole (such as its status, its richness or poverty, its skin colour or its
ability to reach its aims) achieve most of their significance in relation to
perceived differences from other groups and the value connotations of these
differences” (1981, p. 258).
Obviously, the way the in-group is evaluated has important implications for
an individual’s social identity. In other words, one’s own social identity may
be positive or negative depending on the evaluations of those groups that
form one’s social identity. This notion leads to the central hypothesis of SIT:
If it is assumed that individuals strive for a positive self-concept in order to
maintain or enhance their self-esteem, “the in-group must be perceived as
10 Introduction
positively different or distinct from the relevant out-groups” (Tajfel & Turner,
1986, p. 16). Thus, SIT hypothesises that in every intergroup context people
are driven both to gain a positive image and to make the context as
unambiguous and meaningful as possible; they try to satisfy both needs by
means of a positive differentiation of the in-group from the out-group on
those dimensions which are relevant for the comparison.
It is through this hypothesis that the results of the minimal group experi-
ments can be explained. Since the context created by the researchers was
deliberately vague and ill-defined, subjects looked for some sort of meaning
and structure by using the only categorisation available, that is, the one based
on the distinction between Klee and Kandinsky supporters. The categorisa-
tion process makes the two groups cognitively central and clearly distinct
from each other. A positive identity is then achieved through a social com-
parison based on the only available dimension of comparison, that is, the
distribution of money.
However, in the real world the achievement and maintenance of a positive
social identity is a much more complex matter. Rather than being populated
by such groups as Klee and Kandinsky supporters, the social world comprises
groups and categories that stand in status relations with one another, and are
often in competition for resources, prestige, and power. This means that,
while members of more powerful groups normally have a positive social
identity, subordinate groups may confer negative social identity upon their
members, especially if the values of the dominant groups are accepted. As a
consequence, SIT hypothesises that powerful groups try to maintain their
positive identity by maintaining the status quo, while subordinate groups may
either accept the superiority of the more powerful group, which may be
associated with out-group favouritism (Reynolds, Turner, & Haslam, 2000),
or use a variety of strategies in order to improve their social identity.
The strategy chosen usually depends on individuals’ system of belief.
Tajfel makes a distinction between two main systems of belief, namely “social
mobility” and “social change”. The system of beliefs in social mobility holds
that boundaries between social groups are flexible and permeable, and that
individuals can move from one group to another through talent and effort.
Those individuals who subscribe to this system of belief usually try to achieve
a positive identity by means of an individualistic strategy defined as “indi-
vidual mobility”, consisting of leaving their group in order to enter a higher-
status group. On the other hand, social change beliefs make the supposition
that group boundaries are impermeable and that it is therefore impossible to
leave one’s group. In this case social identity may be improved through the use
of one of two general strategies, depending on the perception of the degree of
legitimacy of the status quo: Where the status quo is seen as essentially legit-
imate, subordinate groups may employ some form of “social creativity”
aimed at re-evaluating the group (“We may not be as smart as them, but we’re
better in the sense of being nicer and more honest”). On the other hand, where
the status quo is perceived as illegitimate, subordinate groups may tend to
 Bennett and Sani 11
choose the strategy of “social competition”, in which real attempts are made
to change the relative positions of the in-group and the out-group on key
dimensions (e.g., in industrial settings, in union-based action).

Self-categorisation theory
A recognised weakness of SIT, rather ironically perhaps, was that it had little
to say about the cognitive aspects of social identity salience. Moreover, the
focus of SIT was intergroup processes; intragroup processes, for the most
part, were ignored. Self-categorisation theory (SCT), pioneered by John
Turner and colleagues (Oakes, Haslam, & Turner, 1994; Turner et al. 1987),
sought to address these weaknesses.
SCT shares with SIT the crucial postulate that in many circumstances
individuals define themselves in terms of their group membership, and in
turn a group-defined perception of the self produces specific psychological
effects that affect social behaviour. However, the theory differs from SIT in
that it has a greater explanatory scope, as it is not limited to issues of social
structure and intergroup relations.
As a general social psychological theory of group phenomena, SCT was
explicitly created in order to deal with the individual–group dilemma, which,
according to Turner, is to be conceptualised in terms of the following ques-
tions: “How does a collection of individuals become a social and psycho-
logical group? How do they come to perceive and define themselves and act as
a single unit, feeling, thinking and self-aware as a collective entity?” (Turner,
Hogg, Oakes, Reicher, & Wetherell, 1987, p. 1). The dominant theories of the
1960s reduced the group to little more than a complex system of inter-
personal relations based upon the interdependence and attraction among the
group’s members. The concept of “group” was seen as virtually superfluous
since it was considered to be descriptively useful but unnecessary at an
explanatory level. However, SCT emphasises that a collection of people will
act as a group as long as they feel that they belong to the same whole, that is,
to the same social category. This conviction is principally based on the result
of the experiments on “minimal groups” discussed earlier.
According to Turner and his colleagues (Turner et al., 1987), group forma-
tion precedes – rather than follows – most of the interpersonal phenomena
that characterise our daily life. To be more specific, group formation is seen as
“an adaptive social psychological process that makes social cohesion,
cooperation and influence possible” (Turner et al., 1987, p. 40). This way of
conceiving the group makes it possible for the concept of “group” itself to
become essential from both an empirical and a theoretical point of view.
The key aspects of SCT, which we will consider shortly, can be described in
terms of three fundamental aspects of group functioning: (a) the antecedents
of the psychological group; (b) the basic process underlying the psychological
group; and (c) the consequences of the psychological group. First, however,
it is necessary to mention the main assumptions underpinning the theory.
12 Introduction
Central assumptions of self-categorisation theory
The central, even defining, assumption is that in any circumstance the way in
which an individual experiences him/herself depends on the specific self-
conception that is “activated”. In turn, a self-conception derives from a
self-categorisation, which consists of the cognitive grouping of the self as
identical to some class of stimuli in contrast to some other class of stimuli.
SCT then assumes that any specific self-categorisation is part of a hier-
archical system of classification. Self-categorisations form at different levels
of inclusiveness. For instance, a given self-category (e.g., “social science stu-
dent”) is more inclusive than another (e.g., “sociology student”), where the
former can contain the latter, but the latter cannot contain the former.
A further important assumption is that although the number of levels of
abstraction is potentially infinite, each self-category can be seen – for the sake
of theoretical clarity – as pertaining to one of three main levels of abstrac-
tion. The superordinate level is that of the self as a human being, by means
of which the self is grouped as similar to all human beings, and as different
from all other species in nature. When one is identified in a given situation
with the human race, one is said to behave according to human identity. The
intermediate level concerns the self as member of a social group; at this level
the self is grouped as similar to the other members of the in-group and as
different from the members of the out-groups. Categories like “biologist”,
“French”, “liberal”, “Juventus supporter”, “Greenpeace activist”, and so on
belong to this level of abstraction. When self-conception in a specific circum-
stance is shaped by the categorisation as a member of a social group, one is
seen as behaving and self-perceiving according to one’s social identity.
Finally, there is the subordinate level of abstraction, which is that of the self
as an individual person. In this case the person perceives him/herself as
unique and idiosyncratic, that is, as different and distinct from anybody else.
When prominent characteristics are related to one’s specific and particular
features (e.g., kindness, intelligence, etc.), one is seen as behaving according to
personal identity.
This hierarchical system of self-categories is equivalent to all systems of
natural categories as described by Rosch (1978). However, in opposition to
Rosch’s point of view on natural categories, SCT does not propose that some
self-categories are more central, or somehow more important, than others.
There is not a way of categorising, and therefore experiencing, the self, that
can be said to be deeper or more authentic than other ways. Thus, when we
act on the base of our social identity, and therefore structure the context of
action in terms of “we” versus “them”, we are not experiencing the self in a
way that is less valuable than when we act according to our personal identity.
Turner, Oakes, Haslam, and McGarty (1994) explain:

when we think of and perceive ourselves as “we” and “us” (social iden-
tity) as opposed to “I” and “me” (personal identity), this is ordinary and
 Bennett and Sani 13
normal self-experience in which the self is defined in terms of others who
exist outside the individual person doing the experiencing and therefore
cannot be reduced to personal identity. At certain times the self is defined
and experienced as identical, equivalent, or similar to a social class of
people in contrast to some other class. The self can be defined and
experienced subjectively as a social collectivity”.
(pp. 454–455) [italics in the original]

Antecedents of the psychological group

As we mentioned earlier, SCT hypothesises that “psychological group for-
mation takes place to the degree that two or more people come to perceive
and define themselves in terms of some shared in-group–out-group categor-
isation” (Turner et al., 1987, p. 51). This hypothesis implies that the group
with which a collection of people identify in a given situation can be deter-
mined either by the characteristics of a much greater number of people than
those present in that situation, or by the features exclusively shared by those
present. In Hogg’s (1987) words, “five individuals in a room have just as
much become a group in that situation if they are behaving in terms of a
large-scale category membership they share (such as race, sex, or religion) as
if in terms of an emergent group whose norms and defining features are
unique to that specific collection of five people (a friendship group, an
experimental decision-making group, etc.)” (p. 103). This principle is
important because it represents a “rupture” with traditional social psych-
ology of group processes, in the sense that it obviates the usual distinction
between small groups and large-scale social categories as far as the central
psychological mechanisms bounding people together are concerned. Indeed,
this principle takes it that face-to-face interaction is unnecessary for group
formation.
Once it is established that the existence of a shared self-categorisation is a
necessary condition for a collection of people to feel that they are a group,
the question arises of how it is that within a particular context certain self-
categories become salient while other potential self-categories remain in the
background. In dealing with this problem, SCT borrows from the theory of
Bruner (1957) on the activation of categories in the perceptual field. For SCT
the salience of a particular self-category is determined by an interaction
between “accessibility” and “fit.” Accessibility refers to the tendency to use
those categories that are meaningful in terms of past experience and current
expectations, goals, needs, and values. Oakes, Haslam, and Turner (1994)
also use the term “perceiver readiness”, which clearly indicates that certain
categories are more readily used because of their relevance, usefulness, and
centrality, and because they are likely to be confirmed by the evidence of
reality.
Fit refers to the degree of correspondence between the categories selected
and the stimulus reality to be represented by the categories. Fit has two
14 Introduction
aspects: “comparative fit” and “normative fit” (Oakes, 1987; Oakes, Turner,
& Haslam, 1991). Comparative fit is defined by the principle of metacontrast,
according to which “any collection of individuals in a given setting is
more likely to categorise themselves as a group (become a psychological
group) to the degree that the subjectively perceived differences between them
are less than the differences perceived between them and other people
(psychologically) present in the setting (i.e., as the ratio of intergroup to
intragroup differences increases)” (Turner et al., 1987, pp. 51–52).
Normative fit refers to the fact that the objective differences between
groups must also match the expected stereotypical features of the groups
themselves. For example, to categorise a set of people as Liberals as
opposed to Conservatives, they must not only differ from Conservatives in
their attitudes and behaviour more than from one another, but they must
also be different on particular content dimensions of comparison that are
typically seen as differentiating the two categories. This means that the
categorisation in terms of Liberals versus Conservatives would probably take
place promptly in a situation in which, for instance, the Liberals were arguing
in favour of abortion, while the Conservatives were arguing against it.
The main implication of the fit hypothesis is that self-categories become
salient as a result of an intergroup comparison, and that they therefore vary
as a function of the frame of reference. They can vary in different forms. First
of all they vary in level of inclusiveness. A person can categorise him/herself
either as Sicilian, or Italian, or European, usually depending on the extension
of the context. Self-categories also vary in meaning, in order to reflect the
content of the diagnostic intergroup differences in a given situation. For
example, Haslam, Oakes, Turner, and McGarty (1995) found that Australians
perceive themselves as sportsmanlike, straightforward, and happy-go-lucky
when judged on their own, but when judged in the context of Americans they
see themselves as even more sportsmanlike, less happy-go-lucky, but also
pleasure-loving. Finally, self-categories vary in internal structure, that is, in
terms of the relative prototypicality of their members. In other words, cat-
egories are not defined by a fixed prototype: The relative prototypicality of a
member of a category changes according to the context.
The variability and fluidity of self-categories is a core feature of SCT. In
fact, it underlines the functionality of the self-categories. As stressed by
Turner et al. (1994):

the concept of self as a separate mental structure does not seem neces-
sary, because we can assume that any and all cognitive resources – long-
term knowledge, implicit theories, cultural beliefs, social representations,
and so forth – are recruited, used, and deployed when necessary to create
the needed self-category. Rather than a distinction between the activated
self and the stored, inactive self, it is possible to think of the self as the
product of the cognitive system at work, as a functional property of the
cognitive system as a whole” (p. 459) [italics in the original].
 Bennett and Sani 15
Thus, fluidity and variability of self-categories reflect “functionality” and
“adaptation”. Self-categories are created and made salient in order to facili-
tate behaviour that is appropriate to the situation.

The basic process of the psychological group

When a collection of people share the same social identity, that is, when they
categorise themselves as members of the same social group, a core process
called depersonalisation takes place. The notion of depersonalisation refers
to the fact that when a collective self-identification is salient, “individuals
tend to define and see themselves less as differing individual people and
more as the interchangeable representatives of some shared social category
membership” (Turner et al., 1994, p. 455).
Depersonalisation is facilitated by the already-mentioned “accentuation
effect”, by means of which, when the context is divided into discrete cate-
gories, there is a tendency to exaggerate intragroup similarities and inter-
group differences. This promotes among members of the same group a
sense of being similar (identical, equivalent, interchangeable), and leads to
self-perception in terms of “we” and “us” instead of “I” and “me”. This
mechanism also generates a perception of out-group members in terms of
“they” and “them”. For example, when a psychologist categorises herself as a
“psychologist” in contrast to, say, sociologists, she will accentuate per-
ceptually her similarities to other psychologists – reducing at the same time
her specific personal differences from other psychologists – and will magnify
her stereotypical differences from sociologists.
To be depersonalised does not imply the emergence of some sort of primi-
tive or inferior form of identity, but a change in the level of abstraction of the
self-category. There is a shift from personal identity to social identity, which is
cognitively and behaviourally functional in terms of adaptation of the self to
the social context of action. Indeed, in many circumstances (for example,
involving intergroup competition) it is more adaptive to perceive ourselves as
interchangeable members of a category rather than as unique individuals.
That perception and cognition are at the service of the self and its needs,
goals, and values, is hardly surprising. As Oakes, Haslam and Turner note,
“perception that was not relative to self would be pointless, futile, unimagin-
able and meaningless. It would not be human. [. . .] It is hard to imagine
how and why any biological (or other) system would evolve a capacity for
perception unaffected by the functions, aims, properties and location of the
perceiver” (1994, p. 205).
16 Introduction
The consequences of depersonalisation
Depersonalisation is seen by SCT as the basis of all those processes that
typically take place within a group context. Processes such as attrac-
tion, cohesion, ethnocentrism, cooperation, influence, polarisation, crowd
behaviour, and so on can be seen as resulting from depersonalisation, via
the assumption that salient self-categories tend to be positively evaluated, and
that the self and others are positively evaluated to the degree that they are
perceived as prototypical of the self-category.
Consider, for instance, mutual attraction. Depersonalisation increases the
mutually perceived prototypicality of in-group members on the stereotypical
dimensions that characterise the in-group category. As a result, since the
in-group category is positively valued, the mutual attraction between
members of the in-group will increase. Ethnocentrism, that is, attraction to
one’s group as a whole, is explained in a similar fashion: It is the consequence
of the perceived prototypicality of the in-group in comparison with relevant
out-groups on relevant dimensions. With regard to social cooperation, this
is seen as resulting from the fact that the perception of similarity between
oneself and in-group members, induced by depersonalisation, leads to a
perceived commonality of interests, which in turn increases the level of
intragroup cooperation.
According to SCT, crucial social phenomena such as social influence and
conformity can also be interpreted as being determined by depersonalisation
(Turner et al., 1987). The process leading from depersonalisation to con-
formity via social influence has been named “referent informational influ-
ence” and comprises three distinct phases. (1) A collection of people sharing
the same social identity – who are therefore depersonalised – tend to agree,
or at least expect to agree, with each other, as far as their judgment of, and
reactions to, certain stimuli are concerned. (2) This state of affairs leads the
group members either to learn the existing stereotypic norms of the group,
or to create some common norms tout court. Since the norms of a social
group are generally rather flexible and multifaceted, there is usually a sub-
group or an individual that best expresses and embodies those norms. This
subgroup, or individual, can be considered as prototypical, and as such is
seen as the most appropriate referent for information about valid and cor-
rect norms. (3) Once the norms are known, group members assign these
norms to themselves along with other stereotypical characteristics of the
category, and in so doing they conform to the normative behaviour of
the group.
In conclusion, this way of explaining group phenomena reflects SCT’s
contention that the sense of being a group is not a consequence of inter-
dependence and attraction but instead precedes such group phenomena. As
such, it seems fair to judge that SCT fulfils its aim of reinstating the “group”
as a central psychological and theoretical tool in the explanation of collective
phenomena. (For a fuller account of the social identity approach, see
 Bennett and Sani 17
Haslam, 2001; for recent extensions and modifications to the “classic”
position presented here, see Turner & Onorato, 1999.)

ORGANISATION OF THE BOOK

Briefly stated, the first two substantive chapters of this volume address basic
and general issues in the study of children’s social identities. Following this,
several chapters each examine a particular social identity: gender, nationality,
and ethnicity. Next, applications of the social identity perspective to phe-
nomena contingent upon identity are considered. Finally, commentaries
on the book’s chapters are provided. Although the basic structure is to some
extent artificial, in the sense, for example, that all the chapters at some level
address both identities and applications, it is a structure that draws attention
to the distinction between self-conceptions and phenomena that are con-
tingent upon self-conceptions. Looking to the future, we believe that one of
the major contributions of developmentally oriented research inspired by the
social identity perspective will be the articulation of the relationship between
developing self-conceptions and many diverse forms of social action. How-
ever, coming back to the present, we turn to a more detailed account of the
chapter contents.
In any relatively new field of inquiry, such as the focus of the present
volume, it is vital not to lose sight of the relevance of previous research. Such
research, though perhaps not directly pertinent to new issues, may nonethe-
less provide important insights or may offer a basis for the generation of new
hypotheses. On the other hand, conducting work in relatively uncharted terri-
tory poses fundamental challenges of conceptualisation. The volume’s first
substantive chapter tackles these twin demands admirably. Diane Ruble,
Jeannette Alvarez, Meredith Bachman, Jessica Cameron, Andrew Fuligni,
Cynthia Garcia Coll, and Eun Rhee provide a broadly based discussion that
serves to contextualise many of the issues arising in later chapters. They
provide a valuable review of the available research on children’s understand-
ing of and identification with social categories, particulary gender, race, and
ethnicity. Ruble et al. go on to make the case that social identity must be
understood as a complex, multidimensional construct. Specifically, four
dimensions are addressed: salience, centrality, knowledge, and evaluation.
Attention is given to developmental evidence pertaining to each of these
dimensions, and the case is made showing a pressing need to explore these
key components of identity further. Finally, Ruble et al. consider the con-
sequences of developmental changes in collective identity, for example, with
respect to in-group biases, self-esteem, information-search, and personal
choices.
Fabio Sani and Mark Bennett, in the following chapter, address three main
themes bearing upon the development of social identity. First of all, they
outline their work on children’s conceptions of the normative features of
18 Introduction
social groups, arguing that up until mid-childhood, children’s focus is upon
the dispositional and behavioural features defining a social identity. Only at
mid to late childhood do they start to consider the role of socially shared
beliefs associated with many identities. Next, they make the case that not only
are there developmental variations in conceptions of social groups, but there
are also contextual variations. That is, drawing upon SCT they argue that
children’s conceptions of particular groups vary as a function of the context
in which they consider particular groups. Evidence is provided suggesting
that flexibility in conceptions of groups increases with age. Finally, Sani and
Bennett raise questions about the extent to which previous research relying
on verbal self-descriptions has provided insights into the internalisation of
social identities. Using two separate methods, they provide evidence suggest-
ing that the internalisation of social identities may not be established until at
least the age of 7 years.
The chapters that follow each address a particular type of identity. The first
two consider perhaps the most fundamental of social identities, namely
gender. In looking at research on aspects of gender identity, Kimberly
Powlishta notes extensive evidence for the sorts of generic intergroup
processes predicted by the social identity approach. For example, from an
early age, children show strong preferences for members of their own gender,
and accentuate within-group similarities and between-group differences
(although interestingly, these findings seem to be more pronounced in girls
than boys). On the basis of such findings, Powlishta argues that,

Children may view their own sex as superior in order to achieve a positive
social identity, which in turn should motivate them to adopt gender-
typed characteristics and prefer same-sex playmates. The resulting
gender-segregated play exposes boys and girls to different socialization
contexts, potentially creating or amplifying sex differences. The generic
tendency to exaggerate between-group differences and within-group
similarities may contribute to gender stereotypes, further encouraging
boys and girls to adopt different behaviors and roles.

Thus generic intergroup processes, she suggests, play an important role in

gender-role development.
The following chapter, by Barbara David, Diane Grace, and Michelle Ryan,
also looks at gender identity, but where Powlishta’s focus is largely on social
identity theory, David et al. consider self-categorisation theory. As such, they
see the self in process rather than structural terms and argue that “gender is
one aspect of the flexible, changing, context-dependent self process. What it
means to a child to be a girl or a boy will change with context, as it does for
adults, but the changes will be more dramatic due to developmental growth in
cognitive skills and knowledge . . .”. Drawing upon both SCT and cognitive
developmental considerations, they make the novel case that early gender
identity is no more than an aspect of personal identity; only later does it
 Bennett and Sani 19
become a social identity, properly conceived. More generally, in arguing for
SCT, they make the case that the adoption of a more fluid conceptualisation
of self addresses some of the shortcomings of classic approaches to gender
development.
Martyn Barrett, Evanthia Lyons, and Arantza del Valle consider the social
identity approach in the context of national identity. They outline eight key
predictions made by SIT and SCT and examine how far the predictions are
upheld by their own extensive, multinational studies of national identity
development. In contrast to the research reported by Powlishta on gender
identity, Barrett et al. find inconsistent support for the predictions. Indeed,
the only finding to emerge consistently in all national groups (though only on
one of their measures) was preference for the national in-group, which
appears from 6 years of age. The striking feature of their data is the extent to
which different patterns of findings emerge in different national contexts.
Thus, in seeking to make a judgment about the value of the social identity
approach in a developmental context, Barrett et al. argue that, taken at face
value, the approach might be seen as falling short. However, drawing upon
Turner’s more recent writings (e.g., Turner & Onorato, 1999), they propose
that

the extent to which particular identity phenomena are exhibited may

actually be a product of a much more complex interaction between, for
example, the strength of subjective identification with the in-group, the
individual’s beliefs about the nature of the group boundaries (for
example, whether they are legitimate or illegitimate) and about the status
of the different groups within that system (for example, whether the
ingroup is high or low status, and whether this status is secure or
insecure), and the individual’s own personal motivations, values, needs
and expectancies in relationship to these beliefs.

From this, Barrett et al. make two important observations. One is that future
research may need to examine complex patterns of findings rather than
correlations between pairs of variables. However, their more trenchant obser-
vation is that, to the extent that some cognitive or motivational variable
might always be appealed to, post hoc, to “save” the theory, recent formula-
tions of SCT, being invulnerable to refutation, might arguably be viewed as a
degenerating research programme.
Maykel Verkuyten draws upon SIT and SCT to examine ethnic identity in
children. Looking at majority-group Dutch children and Moroccan, Suri-
namese, and Turkish immigrant children in the Netherlands, he demonstrates
that not only are there important differences between the majority group and
the minorities, but there are also differences between ethnic minority groups.
In addition, there are important identity-related individual differences within
ethnic groups (e.g., in collectivism and in experiences with discrimination).
Verkuyten argues that “Ethnic identity does not only depend on cognitive
20 Introduction
structures and processes but is also determined by social beliefs and context
. . . . . . status differences, cultural values and situational conditions play an
important role in children’s ethnic self-understanding”. Although SCT is
seen as an important theoretical resource, Verkuyten comments critically that
its emphasis is on “the processes of category use in context, and not on how
self-understandings are conceptually organized and change in structure with
age”. To this extent, interestingly, Verkuyten’s position is somewhat different
from that of David et al., in that he asserts the need to acknowledge cross-
situationally stable features of the self, and developmental changes in those
features.
The chapters just described each consider a particular type of identity and
look at phenomena predicted by the social identity approach. In contrast, the
chapters that follow take a particular phenomenon and ask how a social
identity approach could cast light on it. Thus, Drew Nesdale considers ethnic
prejudice in children and begins by arguing that traditional approaches
(e.g., cognitive and socialisation-based accounts) are inadequate to account
for the considerable body of evidence on this matter. Instead, he proposes an
identity-based model of the development of children’s prejudice. Nesdale’s
social identity development theory (SIDT) draws both upon social identity
theory and self-categorisation theory and takes it that children’s early inter-
group biases are expressions of in-group preference rather than out-group
hostility (i.e., liking the in-group more than out-groups, but not actively
disliking out-groups). The appearance of subsequent prejudice is seen as
resulting from the incorporation of negative out-group attitudes as part of an
in-group identity. He states that, “Importantly, SIDT emphasises the critical
significance of social identity processes in the development of children’s
ethnic attitudes and, in so doing, facilitates a long overdue shift away from
the prevailing emphasis in much social developmental research on the pre-
dominance of cognitive processes”. Much evidence is provided that lends
support to this position.
Similarly, Adam Rutland’s chapter argues against traditional accounts of
prejudice suggesting that it is an inevitable consequence of young children’s
cognitive limitations. Rutland begins by noting previous research that has
been taken to support cognitive-developmental accounts of prejudice (e.g.,
Aboud, 1988), particularly that which shows a “peaking” of prejudicial
attitudes at around 7 years of age, with a decline thereafter. In a series of
experiments, he explores the possibility that this developmental progression
instead reflects an increasing capacity, with age, to regulate the expression of
prejudice in a socially appropriate way. Consistent with his hypothesis, he
shows that when in a state of heightened self-consciousness, 6- to 8-year-old
children reveal no evidence of prejudicial attitudes (though they did so in a
standard condition). However, perhaps the most novel aspect of Rutland’s
research is the employment of implicit measures of prejudice, that is,
response latency measures (widely used on adults), which tap aspects of cog-
nition that are relatively automatic. Using such techniques, older children’s
 Bennett and Sani 21
out-group attitudes were found to be significantly more negative than was the
case when they were directly questioned about their attitudes. According to
Rutland, children can thus be seen as social tacticians who strategically
express their intergroup attitudes through a process of effortful on-line con-
struction in context. Rutland’s findings are valuable both because they raise
difficult questions for theoretical accounts that explain prejudice in cognitive-
developmental terms, and because they demonstrate the value and tract-
ability of response latency techniques in developmental work.
In a relatively unusual application of the social identity approach, Brenda
Banker, Sam Gaertner, John Dovidio, Missy Houlette, Kelly Johnson, and
Blake Riek consider the stepfamily as a key intergroup context in many
children’s lives. They note that bringing together two separate family groups
into a single structure is a complex and problematic matter. In seeking to
understand the difficulties that so often arise in this context, and the factors
associated with positive outcomes for stepfamilies, Banker et al. appeal to the
common in-group identity model. This model emerged out of the social iden-
tity approach and “proposes that factors that induce members of two groups
to conceive of themselves as members of a common, more inclusive in-group,
reduce intergroup conflict by enabling cognitive and motivational processes
that contribute to pro-in-group favouring biases to be redirected to include
former out-group members”. Consistent with the model, two studies of step-
children’s perceptions and experiences provide strong evidence that seeing the
stepfamily as a single entity is a significant causal link in the achievement of
harmony within stepfamilies. Moreover, insofar as the findings point to a
factor (“conditions of contact”) that contributes importantly to the percep-
tion of a single family unit, the practical implications of this work are
considerable.
In the two closing chapters, Dominic Abrams and Kevin Durkin each
reflect on some of the key issues arising out of this volume. Abrams addresses
three main themes and considers these with respect to each chapter. First, he
notes that most chapters in the book address the cognitive and motivational
elements of self-categorisation and social identity. Abrams questions whether
there are other aspects of children’s group memberships that need to be
examined. In particular, he sees a need for further research that tackles “the
processes by which intergroup relationships are defined, sustained, and given
continued meaning for children”. Second, he remarks that chapters typically
focus either on the role of social context or cognitive development, and
asserts that our understanding would be enhanced by directing greater atten-
tion to the fact that identities presuppose shared meanings that in turn
depend on social interaction between people. Finally, Abrams comments
that research on the development of social identity should be expanded to
embrace some of the key controversies that feature in the adult literature – for
example, the impact of different motivational elements of identity, and the
relationship between communication and social identity.
Durkin considers several major themes, reminding us first that for the
22 Introduction
dependent infant, social groups are a “raw necessity”. The family, as the
fundamental social group, provides experiences that will foster children’s
earliest feelings and cognitions about groups. It is also in the context of
family interactions that children will be guided towards socially significant
identities, such as gender and ethnicity. The diverse range of social identities
that children are afforded, suggests Durkin, may militate against the establish-
ment of a broad overarching developmental framework in that “the develop-
ment of any one [identity] is multidetermined and intersects with other
powerful social processes” so that identities may develop differently. In con-
sidering the forces that shape social identity, Durkin goes so far as to note the
need to acknowledge competition between developmental and social pro-
cesses, something that further complicates our understanding in this domain.
Moreover, he asserts the necessity to give greater prominence to the role of
language in the development of social identities. In addition to addressing
formative processes, Durkin also notes that identities have consequences, in
terms of opportunity, social status, well-being, and so on, and more needs to
be understood about such consequences in childhood. In the final part of his
chapter, Durkin observes that little is known about changing social identities
across the lifespan: “The overwhelming expectation we can derive from the
present volume is that we will need to draw on both developmental and social
psychological perspectives if we are to investigate adequately the longer-term
construction and maintenance of social selves”.

REFERENCES

Aboud, F. E. (1988). Children and prejudice. Oxford: Blackwell.

Abrams, D., Rutland, A., Cameron, L., & Marques, J. M. (2003). The development
of subjective group dynamics: When in-group bias gets specific. British Journal of
Developmental Psychology, 21, 155–176.
Allport, G. W. (1968). The historical background of modern social psychology. In G.
Lindzey & E. Aronson (Eds.), The handbook of social psychology, Vol. 1 (2nd ed.).
Reading, MA: Addison-Wesley.
Baldwin, J. M. (1895). Mental development of the child and the race: Methods and
processes. New York: Macmillan.
Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorization and the
formation of intergroup attitudes in children. Child Development, 68, 530–543.
Billig, M. (1976). Social psychology and intergroup relations. London: Academic Press.
Brandstädter, J., & Lerner, R. M. (Eds.) (1999). Action and self-development. Thou-
sand Oaks, CA: SAGE Publications.
Brehm, S. S., Kassin, S. M., & Gibbons, F. X. (Eds.) (1981). Developmental social
psychology: Theory and research. New York: Oxford University Press.
Brewer, M. B. (1999). The psychology of prejudice: Ingroup love or outgroup hate?
Journal of Social Issues, 55, 429–444.
Brooks-Gunn, J., & Lewis, M. (1979). Social cognition and the acquisition of self. New
York: Plenum Press.
Bruner, J. S. (1957). On perceptual readiness. Psychological Review, 64, 123–152.
 Bennett and Sani 23
Case, R. (1992). The mind’s staircase. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Cooley, C. H. (1902). Human nature and social order. New York: Charles Scribner’s
Sons.
Damon, W., & Hart, D. (1988). Self-understanding in childhood and adolescence. New
York: Cambridge University Press.
Deaux, K., Reid, A., Mizrahi, K., & Cotting, D. (1999). Connecting the person to the
social: The functions of social identification. In T. R. Tyler, R. M. Kramer, & O. P.
John (Eds.), The psychology of the social self. Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Demetriou, A., & Kazi, S. (2001). Self-image and cognitive development: Structure
functions and development of self-evaluation and self-representation in adolescence.
London: Routledge.
Doise, W. (1978). Individuals and groups: Explanations in social psychology.
Cambridge: Cambridge University Press.
Doise, W. (1982). Report on the European Association of Experimental Social
Psychology. European Journal of Social Psychology, 12, 105–111.
Durkin, K. (1995). Developmental social psychology: From infancy to old age. Oxford:
Blackwell.
Eckes, T., & Trautner, H. M. (Eds.) (2000). The developmental social psychology of
gender. Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Eiser, J. R., & Stroebe, W. (1972). Categorization and social judgment: European
Monographs in Social Psychology, no. 3. London: Academic Press.
Elms, A. C. (1975). The crisis of confidence in social psychology. American
Psychologist, 30, 967–976.
Fenigstein, A., Scheier, M. F., & Buss, A. A. (1975). Public and private self conscious-
ness: Assessment and theory. Journal of Consulting and Clinical Psychology, 43,
522–527.
Ferguson, C. K., & Kelley, H. H. (1964). Significant factors in over-evaluation of own
groups’ products. Journal of Abnormal and Social Psychology, 69, 223–228.
Fischer, K. W. (1980). A theory of cognitive development: The control and
construction of hierarchies of social skills. Psychological Review, 87, 477–531.
Flavell, J. H., & Ross, L. (1981). Social cognitive development. Cambridge: Cambridge
University Press.
Gergen, K. J. (1973). Social psychology and history. Journal of Personality and Social
Psychology, 26, 309–320.
Harter, S. (1987). The determinants and mediational role of global self worth in
children. In N. Eisenberg (Ed.), Contemporary issues in developmental psychology
(pp. 219–242). New York: John Wiley & Sons.
Harter, S. (1998). The development of self-representations. In W. Damon (Series Ed.)
& N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Social, emotional,
and personality development (5th ed., pp. 553–617). New York: John Wiley & Sons.
Harter, S. (1999). The construction of the self: A developmental perspective. New York:
Guilford Press.
Haslam, S. A., Oakes, P. J., Turner, J. C., & McGarty, C. (1995). Social categorization
and group homogeneity: Changes in the perceived applicability of stereotype
content as a function of comparative context and trait favourableness. British
Journal of Social Psychology, 34, 139–160.
Haslam, S. A. (2001). Psychology in organizations: The social identity approach.
London: Sage.
24 Introduction
Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological
Review, 94, 319–340.
Hogg, M. A. (1987). Social identity and group cohesiveness. In J. C. Turner, M. A.
Hogg, P. J. Oakes, S. D. Reicher, & M. S. Wetherell (1987), Rediscovering the social
group: A self-categorization theory. Oxford: Blackwell.
Hogg, M. A., & Abrams, D. (1988). Social identifications: A social psychology of
intergroup relations and group processes. London: Routledge.
James, W. (1890). Principles of psychology. Chicago: Encyclopaedia Britannica.
Jaspars, J. M. F. (1980). The coming of age of social psychology in Europe. European
Journal of Social Psychology, 10, 421–429.
Jones, E. E., & Berglas, S. (1978). Control of attributions about the self through self-
handicapping strategies: The appeal of alcohol and the role of under achievement.
Personality and Social Psychology Bulletin, 4, 200–206.
Keller, A., Ford, L., & Meacham, J. (1978). Dimensions of self-concept in preschool
children. Developmental Psychology, 14, 483–489.
Kihlstrom, J. F., & Cantor, N. (1984). Mental representations of the self. In
L. Berkowitz (Ed.), Advances in experimental social psychology, Vol. 17 (pp. 2–40).
New York: Academic Press.
Lerner, R. M. (1998). Theories of human development: Contemporary perspectives.
In R. M. Lerner (Ed.), Handbook of child development Vol. 1 (pp. 1–24). New York:
Wiley.
Lewis, M., & Brookes-Gunn, J. (1979). Social cognition and the acquisition of the self.
New York: Plenum Press.
Linville, P. W. (1985). Self-complexity and affective extremity: Don’t put all your eggs
in one cognitive basket. Social Cognition, 3, 94–120.
Markus, H. (1977). Self-schemata and processing information about the self. Journal
of Personality and Social Psychology, 35, 63–78.
Masters, J. C., & Yarkin-Levin, K. (Ed.) (1984). Boundary areas in social and
developmental psychology. New York: Academic Press.
McGarty, C., & Haslam, S. A. (1997). Introduction and a short history of social
psychology. In C. McGarty & S. A. Haslam (Eds.), The message of social
psychology. Oxford: Blackwell.
McGuire, W. (1973). The yin and yang of progress in social psychology: Seven koan.
Journal of Personality and Social Psychology, 26, 446–456.
Mead, G. H. (1934). Mind, self, and society from the standpoint of a social behaviourist.
Chicago: University of Chicago Press.
Moore, C., & Lemmon, K. (Eds.) (2001). The self in time. Mahwah, NJ: Lawrence
Erlbaum Associates, Inc.
Oakes, P. J. (1987). The salience of social categories. In J. C. Turner, M. A. Hogg, P. J.
Oakes, S. D. Reicher, & M. S. Wetherell (Eds.), Rediscovering the social group: A
self-categorization theory. Oxford: Blackwell.
Oakes, P. J., Haslam, A., & Turner, J. C. (1994). Stereotyping and social reality.
Oxford: Blackwell.
Oakes, P. J., Turner, J. C., & Haslam, S. A. (1991). Perceiving people as group
members: The role of fit in the salience of social categorizations. British Journal of
Social Psychology, 30, 124–144.
Pomerantz, E. M., & Newman, L. S. (2000). Looking in on the children: Using
developmental psychology as a tool for hypothesis testing and model building in
social psychology. Personality and Social Psychology Review, 4, 300–316.
 Bennett and Sani 25
Rabbie, J. M., & Horwitz, M. (1969). Arousal of ingroup–outgroup bias by a chance
win or loss. Journal of Personality and Social Psychology, 13, 269–277.
Rabbie, J. M., & Wilkens, G. (1971). Ingroup competition and its effect on intragroup
relations. European Journal of Social Psychology, 1, 215–234.
Reicher, S. D. (1984). Social influence and the crowd: Attitudinal and behavioural
effects of deindividuation in conditions of high and low group salience. British
Journal of Social Psychology, 23, 341–350.
Reynolds, K. J., Turner, J. C., & Haslam, S. A. (2000). When are we better than them
and they worse than us? A closer look at social discrimination in positive and
negative domains. Journal of Personality and Social Psychology, 78, 64–80.
Ring, K. (1967). Experimental social psychology: Some sober questions about some
frivolous values. Journal of Experimental Social Psychology, 3, 113–123.
Rochat, P. (2001). Origins of self-concept. In G. Bremner & A. Fogel (Eds.), Blackwell
handbook of infant development (pp. 191–212). Oxford: Blackwell.
Rosch, E. (1978). Principles of categorization. In E. Roschand & B. B. Lloyd (Eds.),
Cognition and categorization. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Schlenker, B. (1980). Impression management: The self-concept, social identity, and
interpersonal relations. Monterey, CA: Brooks/Cole.
Sears, D. O. (1986). College sophomores in the laboratory: Influences of a narrow
databasee on social psychology’s view of human nature. Journal of Personality and
Social Psychology, 51, 515–530.
Sherif, M. (1967). Group conflict and co-operation: Their social psychology. London:
Routledge & Kegan Paul.
Snyder, M. (1974). Individual differences in self-control of expressive behaviour.
Dissertation Abstracts International, 33, 4533A–4534A.
Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of
the self. In L. Berkowitz (Ed.), Advances in experimental social psychology, Vol. 21
(pp. 261–302). San Diego, CA: Academic Press.
Swann, W. B., & Read, S. J. (1981). Self-verification processes: How we sustain our
self-conceptions. Journal of Experimental Social Psychology, 17, 351–370.
Tajfel, H. (1970). Experiments in intergroup discrimination. Scientific American, 223,
96–102.
Tajfel, H. (1972). Social categorization. In S. Moscovici (Ed.), Introduction a la
psychologie sociale, Vol. 1. Paris: Larouse.
Tajfel, H. (1978). Differentiation between social groups: Studies in the social psychology
of intergroup relations. London: Academic Press.
Tajfel, H. (1981). Human groups and social categories. Cambridge: Cambridge
University Press.
Tajfel, H. (1984). Intergroup relations, social myths and social justice in psychology.
In H. Tajfel (Ed.), The social dimension, Vol. 2. Cambridge: Cambridge University
Press.
Tajfel, H., Flament, C., Billig, M. G., & Bundy, R. F. (1971). Social categorization and
intergroup behaviour. European Journal of Social Psychology, 1, 149–177.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G.
Austin & S. Worchel (Eds.), The social psychology of intergroup relations. Monterey,
CA: Brooks/Cole.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behaviour.
In S. Worchel & W. G. Austin (Eds.), Psychology of intergroup relations (2nd ed.).
Chicago: Nelson-Hall.
26 Introduction
Tajfel, H., & Wilkes, A. L. (1963). Classification and quantitative judgement. British
Journal of Psychology, 54, 101–114.
Tesser, A. (1988). Toward a self-evaluation maintenance model of social behaviour.
In L. Berkowitz (Ed.), Advances in experimental social psychology, Vol. 21
(pp. 181–227). New York: Academic Press.
Turner, J. C. (1975). Social comparisons and social identity: Some prospects for
intergroup behaviour. European Journal of Social Psychology, 5, 149–178.
Turner, J. C. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel
(Ed.), Social identity and intergroup relations. Cambridge: Cambridge University
Press.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorization theory. Oxford: Blackwell.
Turner, J. C., Oakes, P. J., Haslam, S. A., & McGarty, C. A. (1994). Self and collective:
Cognition and social context. Personality and Social Psychology Bulletin, 20,
454–463.
Turner, J. C., & Onorato, R. S. (1999). Social identity, personality, and the self-
concept: A self-categorization perspective. In T. R. Tylor, R. M. Kramer, & O. P.
John (Eds.), The psychology of the social self (pp. 11–46). Mahwah, NJ & London:
Lawrence Erlbaum Associates, Inc.
Turner, J. C., & Reynolds, K. J. (2001). The social identity perspective in intergroup
relations: Theories, themes, and controversies. In R. Brown & S. Gaertner (Eds.),
Blackwell handbook of social psychology: Intergroup processes (pp. 133–152).
Oxford: Blackwell.
Wills, T. A. (1981). Downward comparison principles in social psychology.
Psychological Bulletin, 90, 245–271.
Part I

Basic issues
2 The development of a sense of
“we”: The emergence and
implications of children’s
collective identity
Diane N. Ruble, Jeannette Alvarez,
Meredith Bachman, Jessica Cameron,
Andrew Fuligni, Cynthia Garcia Coll,
and Eun Rhee

The emergence of the self is one of the most fundamental developmental

tasks faced by the young child (see Durkin, 1995; Kohlberg, 1969; Lewis,
1990). Self-concept development begins in early childhood and continues
through adolescence and adulthood. This self-concept consists of both a
personal self (a sense of self as different from the other) and a social self (a
sense of self as connected to the other) (see Brewer & Gardner, 1996). This
latter social self consists of not only the selves present when interacting with
family and friends, but also with the selves that are a product of the child’s
membership in particular social groups.
In understanding the development of the self, two primary perspectives
have dominated: (1) assessing the structure and content of the personal
self-concept at different developmental periods (e.g., Damon & Hart, 1988;
Harter, 1983; Livesley & Bromley, 1973) and (2) examining the role of signifi-
cant others, specifically parents, on the development of the self (e.g., Bowlby,
1980; Cooley, 1902; Higgins, 1991; Mead, 1934). Clearly absent from these
approaches is the development of the self that is constructed on the basis of
one’s membership in particular social categories (e.g., gender, ethnicity/race).
Indeed, in Harter’s (1998) chapter on the self in the Handbook of child
psychology, the only mention of social identity is a brief note of the relation
between ethnicity and self-esteem.
This is not to say that psychologists have not assessed the development of
specific social categories, primarily gender and race, in children. In particular,
there has been a focus on the developmental stages necessary for the
achievement of gender and racial constancy: identification of the gender/race
of self and other, the stability of gender and race over time, and the immut-
able nature of gender and race (e.g., Aboud & Skerry, 1983, Rhee, Ruble,
Awarcz, Stangor, & Jones, 2003b; Ruble & Martin, 1998). This approach has
been based on cognitive-developmental theories proposing that understand-
ing the immutability of one’s social identities motivates children to learn the
30 Children’s collective identity
relevant norms and act in accordance with them (Kohlberg, 1966). But this
perspective is limited to a focus on knowledge and awareness of social
categories or labelling oneself as a member of one. What is missing is an
account of real identification – a sense of belonging with a particular social
group. Hence, this perspective does not address such questions as whether a
particular social identity is evaluated positively or negatively, whether it is
salient and central to the self-concept, how it influences important develop-
mental outcomes, such as self-esteem and academic engagement, or how it
changes as a function of context.
The goal of this chapter is to examine the development and consequences
of children’s understanding of and identification with social categories. First,
we begin with a brief description of previous approaches to social identity in
adults and adolescents to provide a theoretical background. Second, we
introduce some of the developmental research on the self and highlight
why social identification is integral to an understanding of changes in child-
ren’s self-knowledge. Third, we discuss the current literature on children’s
developing understanding of social categories and their membership of them.
Fourth, we present social-psychological perspectives on the multidimension-
ality of social identity and review relevant developmental research. Finally,
we outline the importance of taking this social psychological approach in
understanding the link between social identity, the self-concept, and various
developmental outcomes.

THE NATURE OF SOCIAL IDENTITY

Theoretical distinctions
Although the nature and consequences of identification with social groups
has not been a focus of much developmental research, such questions have
been extensively explored by social psychologists. An individual’s social iden-
tities have been defined in different ways, but all definitions refer to certain
key elements. First, they refer to aspects of the self-concept that are defined in
terms of or in relation to other people and groups (e.g., a daughter; a girl)
(Ashmore, Jussim, Wilder, & Heppen, 2001). Second, they are socially con-
structed and interpersonally significant categories (Thoits & Virshup, 1997).
Third, certain values and emotional significance are attached to these role or
category memberships (Tajfel, 1978).
A detailed discussion of this literature is beyond the scope of this chapter
(see Ashmore et al., 2001; Deaux & Martin, 2001, for reviews), but certain
constructs are of particular interest for a developmental analysis. One is a
distinction within social identity between self as fulfilling a role (e.g., teacher)
and self as a member of a collective group (e.g., Asian-American) (Thoits &
Virshup, 1997). Although both types of social identity are relevant to child-
hood, collective identity provides a connection to the long-standing interest
in the development of categorisation in the cognitive-development literature
 Ruble et al. 31
(e.g., Gelman & Markman, 1986) as well as representing influential processes
during early socialisation (e.g., becoming a girl). Thus, in the present chapter,
consistent with this volume’s aims, we focus on the development of children’s
social identities as members of social categories: i.e., collective identity, or
what Thoits and Virshup (1997) have labelled the “we”.
Collective identities are viewed as serving a number of different
motives and having significant personal and interpersonal consequences (see
Hewstone, Rubin, & Willis, 2002; and Brewer & Brown, 1998, for reviews). A
few of these include feelings of inclusion and a need for differentiation, both
of which can be satisfied through identification with optimally distinct groups
(Brewer, 1991) and a reduction of uncertainty through social comparison
processes and by belonging to groups that provide clear norms (Hogg, 2000).
Perhaps most influential in the literature is social identity theory (Tajfel &
Turner, 1979), which proposes that individuals strive for positive self-concept,
and that identification with and comparison across groups are in the service
of this goal. According to Turner (1982), once the self is perceived in terms of
group membership the stereotypes that apply to the group such as status,
prestige, emotional experiences, goals, norms, and traits will all influence
self-perception. The nature of such categorisation and comparison processes
are likely to change dramatically between the early and middle elementary
school years (Lutz & Ruble, 1995; Ruble & Frey, 1991). Moreover, the mere
act of categorising individuals into social groups changes the nature of
interpersonal perceptions and behaviours (Tajfel, 1978). For example, there
is an increased perception of between-group differences and within-group
similarity and increased in-group favouritism. Such “minimal group”
changes occur even if the social categories are completely arbitrary (e.g.,
wearing different coloured teeshirts) (Bigler, Jones, & Lobliner, 1997). For
not so minimal social categorisation (i.e., membership in real social groups),
consequences can be more critical. In the present chapter, we focus on the
development and consequences of collective identity for gender, ethnicity,
and race.

Previous approaches to identity development

Social psychologists have also been interested in some developmental aspects
of social identification, particularly the process by which adolescents and
adults become committed to a particular identity. According to Erikson’s
(1963) theory of ego identity formation, the development of one’s identity
arises out of the identity crisis that takes place during the adolescent years. To
deal with this fundamental psychosocial crisis, one must engage in a period of
exploration of different identities and roles, eventually emerging to commit to
specific occupational, political, and ideological identities (Marcia, 1988).
Perhaps the most concerted attempt to take a developmental approach to
collective identity has taken place within the study of ethnic identity devel-
opment during adolescence. For example, Cross’s model of nigrescence
32 Children’s collective identity
(1978), which was based upon the emergence of race-consciousness among
black college students during the American civil rights period, included a
stage-like progression that began with a period in which one’s ethnic identity
was never explored (“pre-encounter”). This period was then interrupted by
an “encounter”, which represented an awakening of the significance of one’s
race in the social world, followed by an “immersion” stage, which involved
extensive exploration of one’s race and culture. The ultimate end-point of
immersion was internalisation, which was a fuller understanding and
acceptance of one’s racial and cultural identity.
Similarly, Phinney (1989) outlined a three-stage Eriksonian-type progres-
sion that began with an unexamined ethnic identity, followed by an ethnic
identity search and exploration, which eventually culminated in an achieved
ethnic identity. Though very similar to Cross’s model, Phinney’s model did
not include an encounter event that stimulated an exploration or immersion
phase, primarily because Phinney developed a model with several different
ethnic groups in mind.
Although both models implied a developmental progression through
stages, they have been examined mostly from an individual-differences
perspective. That is, adolescents or young adults have been categorised
according to their stage or status, and subsequently compared in terms of
personality traits and well-being (Phinney, 1990). Rarely have the stages of
ethnic identity development been examined within the same individuals over
time in order to test whether there is indeed a typical ordering of stages.
Although some evidence exists to suggest that older adolescents are more
likely to reach “higher” levels of identity development than younger adoles-
cents, it is unclear as to the order at which these youths progressed through
the hypothesised stages (Phinney & Chavira, 1992).
The extent to which an encounter is necessary for ethnic or racial identity
development is also unclear. Such a possibility has been acknowledged
recently by Cross and Fhagen-Smith (2001), who suggest that the majority of
African-American children do develop some type of black identity as a result
of their socialisation experiences inside and outside of the family, implying
that an encounter phase may not be essential. Moreover, both Cross and
Phinney have recently suggested that the final stages of their original models
are not developmental end states, and that individuals may “revisit” earlier
stages and continue to explore the meaning of their ethnic and racial identity
throughout the lifespan (Cross & Fhagen-Smith, 2001; Phinney, 1996). It will
be interesting in future research to examine how collective context (e.g., the
quality of interracial relationships, or being of majority or minority status)
relates to the emergence or re-emergence of these stages.
 Ruble et al. 33
WHEN AND HOW DO CHILDREN UNDERSTAND AND
IDENTIFY WITH SOCIAL CATEGORIES?

The fact that previous research on the development of social identification

has focused primarily on adolescence and adulthood, rather than child-
hood, is surprising in two ways. First, the developmental literature clearly
demonstrates that social categories, particularly gender and race, are salient
and meaningful categories for young children (see, e.g., Aboud, 1988; Ruble
& Martin, 1998). Second, the acquisition of self as distinct from others,
yet connected to others, is a primary goal of social development (Ruble &
Goodnow, 1998). Thus, from infancy until adolescence, individuals are for-
mulating a sense of identity based both on their characteristics, abilities, and
preferences (psychological self) and on their social identification with parents,
peers, and social groups (social self).
Yet the personal self, rather than the social self, has been the primary
emphasis of developmental researchers. Of particular interest have been the
developmental changes in the self-concept that occur between early child-
hood and middle childhood. Prior to the age of 7–8 years, children’s self-
descriptions typically refer to external characteristics and social relationships
(e.g., “I am a boy, my name is Jason. I live with my father and mother in a big
house”, Harter, 1998, p. 47). By the age of 8 years, children’s self-descriptions
consist of statements that refer more to stable trait-like characteristics
(Livesley & Bromley, 1973; Newman & Ruble, 1988) and evaluative state-
ments that involve comparisons with others (Ruble & Frey, 1991). Although
children’s self-descriptions clearly suggest a developmental progression from
the concrete to the psychological, experimental research suggests that young
children can use psychological terms with reference to self and others (e.g.,
Bretherton, McNew, & Beeghly-Smith, 1981; Wellman, 1990) and show an
understanding of global evaluative personal characteristics, such as good or
mean (Eder, 1990; Ruble & Dweck, 1995). Nevertheless, it is not until middle
childhood that persons, including the self, are viewed as having a more gen-
eral, stable, organising structure that directs behaviours (Rholes, Newman, &
Ruble, 1990). Moreover, children’s sense of the stability and predictability of
people is no longer conceived primarily through global evaluations (good
or bad person), but also through recognition of individual differences in
dispositional characteristics (Alvarez, Ruble, & Bolger, 2001).
These changes in self and other perceptions during early to middle child-
hood should have important implications for both children’s understanding
of social categories and their identification with them. First, social categories,
particularly gender and race, should no longer be seen as referring to overt
physical characteristics (e.g., long hair, dark skin), but should now begin to
take on deeper meanings. Second, group membership becomes a comparative
process, whereby information about social categories received from the social
environment can be used to imbue social categories with evaluative meaning
(see J. A. Cameron, Alvarez, Ruble, & Fuligni, 2001). Hence, collective
34 Children’s collective identity
identities may become quite significant to children’s self-concept, at least by
middle childhood, particularly in an environment where a social category
membership is salient (e.g., only black child in a class) or where a social
category is made central (e.g., parents teaching their child about what it
means to be Mexican).
A recent ethnographic study conducted by Connolly (1998) supports our
claim that social identity development must be examined in young children.
Finding fault with the lack of social scientific research that was focused on
young children’s conceptualisations of their social worlds, Connolly spent a
year observing and conducting unstructured interviews with 5- and 6-year-
old children in an English multi-ethnic, inner-city primary school. Based on
this research, Connolly argues that many young children play an active role in
negotiating their identities through drawing upon discourses on race, class,
gender, and age. Connolly’s work is admittedly ethnographic and provides a
description of just one specific community. However, transcripts from his
studies highlight the manner in which very young children often do spon-
taneously discuss race, gender, and other social groupings in a sophisticated
manner. This leads Connolly to question the appropriateness of making
assumptions about children’s cognitive ability or their level of awareness on
matters of race, gender, or sexuality simply by their age. To us, this work
signals the need for a comprehensive exploration of social identity formation
in young children employing developmental psychological methods.
In short, ignoring collective identity development in younger children may
be a serious omission. Middle childhood (ages 7–10) would seem to be a key
period in which to examine the incorporation of social category memberships
into the emerging self-concept. It is during this period that children’s sense of
self becomes more complex as they are exposed to a new social environment
outside of the family. As such this is one of the most important social transi-
tions for development (see Ruble, 1994; Ruble & Seidman, 1996). At school,
children must learn the specific roles, rules, and expectations associated with
peer and teacher relations (e.g., Akiba, Garcia Coll, & Magnuson, 2001,
Higgins & Parsons, 1983). It is through these interactions with peers and
teachers that children develop new social roles (e.g., student) and collective
identities (e.g., Asian), revise current social identities beyond that of the
family, adopt or contest the prevalent social valences associated with each
category, and integrate all of these identities into a coherent self-concept.
Despite the relative lack of attention to social or collective identity in
reviews on the development of self, several sets of literature provide informa-
tion about certain aspects of children’s understanding of and identification
with social categories. In this section, we review those literatures and attempt
to integrate them around the question of the developmental course of
children’s sense of “we”.
 Ruble et al. 35
The development of social category knowledge
and identification
The major social categories relevant to young children are those that involve
clear physical differences – sex and race – suggesting perhaps that children’s
early understanding of social categories begins with an awareness of physical
differences and similarities. Although one might surmise that age would also
be a salient social identity, it has not attracted research interest. There is an
extensive body of literature examining infants’ and young children’s under-
standing of categorical sex and a comparable literature on slightly older chil-
dren’s understanding of categorical race. These findings are reviewed with
respect to three basic levels of children’s developing understanding—
awareness, identification, and constancy – integrating findings of studies of
sex and of race when possible.

Awareness
The term “awareness” is often used very generally to refer to any behaviour
suggesting a response on the basis of sex or race, such as preferences and
attitudes (e.g., Brand, Ruiz, & Padilla, 1974; Newman, Liss, & Sherman,
1983). In this section, we use the term much more specifically to refer only to
children’s ability to distinguish the sexes or races or to label them correctly,
regardless of whether they view themselves correctly as a member of one or
prefer one or the other.
Until recently, studies of awareness have examined children’s ability to
point to or sort photographs in response to social category labels or to gener-
ate the correct label. Studies of gender categories (see Ruble & Martin, 1998,
for a review) have shown that many children begin to label the sexes shortly
after their second birthdays, at least half can do so by 2½ years of age, and
virtually all can sort photographs on the basis of sex by 3 years of age
(Leinbach & Fagot, 1986; Thompson, 1975; Weinraub, Clemens, Sockloff,
Ethridge, Gracely, & Myers, 1984). Interestingly, the ability to label adult
males and females occurs somewhat earlier than the ability to label children
(Leinbach & Fagot, 1986). Children’s ability to label the races seems to occur
a bit later, after 3 years of age (Katz & Kofkin, 1997), but awareness of
ethnicity and racial categories seems to develop rapidly after 4 years of age
(Aboud & Amato, 2001).
Recent studies using habituation techniques have suggested, however, that
infants are capable of discriminating social categories at some level by 12
months of age (Leinbach & Fagot, 1993) and perhaps even younger. For
example, in one study, after viewing repeated instances of faces of a single sex
or race, the face of an unseen sex or race category was found to elicit more
looking in 6-month-olds than a novel but same sex or race face (Katz &
Kofkin, 1997). Interestingly, these researchers found that African-American
infants were more sensitive to racial cues than were European-American
36 Children’s collective identity
infants. Such findings suggest that social category awareness may emerge
before preschool age, but it is important at this point to be cautious. Not
all studies find such awareness of race so young (e.g., Fagan & Singer, 1979),
and the level of social category understanding required for perceptual
discrimination of faces may be quite minimal.

Identification
When can children verbally label themselves accurately in terms of social
categories? The findings suggest that children can label their own sex and
place a picture of themselves with other same-sex children by approximately
24–36 months of age, but there is considerable variability across children and
studies and this ability does not necessarily show a parallel to the ability to
label others (Ruble & Martin, 1998). For example, one recent study found
that the majority of children at 30 months of age were able to identify them-
selves using gender labels but were unable to do so for others (Katz & Kofkin,
1997). Research with infants has not yet tackled directly the question of how
early young children show even implicit understanding of their own category
membership. It is known that children develop a concept of themselves as a
distinct physical entity by 18 months of age and that this achievement pre-
cedes other self-related developments, such as the use of self-referential terms
(Stipek, Gralinski, & Kopp, 1990). Moreover, some research suggests that
infants’ knowledge of attributes associated with gender categories increases
substantially during the second year of life (see Martin, Ruble, & Szkrybalo,
2002).
Studies of children’s racial or ethnic labelling and identification have not
typically looked at very young children. Some studies suggest that children’s
abilities to self-identify using racial labels emerges soon after the awareness of
categorical sex, between the ages of 4 and 5 (Ramsey, 1991; Rhee & Ruble,
1997). In one recent study examining the emergence of sex and racial identifi-
cation in the same children, Katz and Kofkin (1997) reported that by 3 years
of age, most (77%) of the Euro-American children accurately self-labelled
based on race, but only 32% of the African-American children did so. The
authors argue that this difference is not a cognitive one, because the African-
American children performed equally well on gender labelling and had no
difficulty performing a racial sorting task that did not require the use of a
verbal label. Instead, they suggest that race labels may produce more conflict
in African-American children.
Most studies have examined children’s ability to discriminate between
children with black and white skin in both self- and other-identification, but
some studies of ethnic identification have shown similar effects. For example,
Vaughan (1963) reported that Maori and Pakeha children could classify
themselves correctly by age 4 years. There appears to be considerable vari-
ability across studies, however, as to when ethnic identity is learnt, perhaps
because ethnic awareness may require an understanding of relatively complex
 Ruble et al. 37
features such as customs and beliefs (Aboud, 1984; Bernal & Knight, 1997).
Taken together, these various findings suggest that the ability to label oneself
in terms of race or ethnicity may emerge between 3 and 5 years of age, but
other factors determine whether or not children do accurately self-identify at
this age. Moreover, it is not clear whether, at this age, such identification
means anything more than empty labels (e.g., “I’m Mexican because my
mother said so”; Bernal & Knight, 1993, p. 35).

Constancy
Kohlberg (1966) argued that when children learn that a social category is
not changeable, they have achieved a new level of understanding. According
to his cognitive-developmental model, children’s comprehension of social
category membership changes in accordance with cognitive development,
specifically the acquisition of conservation. He proposed three stages leading
to children’s comprehension of the immutability of sex category member-
ship: (1) accurate identification of category membership for oneself and
others; (2) stability of category membership over time; and (3) consistency of
category membership across superficial transformations in appearance or
context. According to Kohlberg, the development of this understanding was
highly significant, because once children understood the unchanging nature
of being a boy or a girl, mastery motives propel them to seek out information
about what is appropriate for their sex and to behave in accordance with the
conclusions that they draw (Stangor & Ruble, 1987). Similar motivational
processes would seem to apply to racial constancy, and a few researchers
have examined this hypothesis (e.g., Bernal, Knight, Garza, Ocampo, & Cota,
1990; Ocampo, Bernal, & Knight, 1993; Semaj, 1980).
The literature generally suggests that gender constancy develops in three
stages (identity, stability, and consistency) between 3 and 7 years of age.
There is some controversy about the exact developmental time course, how-
ever, due largely to methodological issues. For example, the attainment of
gender constancy can vary depending on whether children are provided only
with a forced-choice measure (e.g., requiring only a yes or no response) or are
additionally asked to explain their choices (Ruble & Martin, 1998). Although
many 3- to 4-year-old children give mostly correct responses to forced-choice
questions about whether, for example, a boy would still be a boy if he
had long hair or wore a dress, it is not until 6–7 years of age that most
children can explain their responses in a way that indicates true constancy
understanding (Szkrybalo & Ruble, 1999).
Relatively few studies have examined the development of racial or ethnic
constancy. The findings have suggested that this kind of constancy develops
later than gender constancy, sometime between 7 and 9 years of age in white
children (Aboud & Ruble, 1987; Ocampo et al., 1993). The exact develop-
mental trajectory varies considerably, depending on whether race or ethnicity
is being examined and depending on the race or ethnicity of the children
38 Children’s collective identity
participating in the study (Bernal et al., 1990; Semaj, 1980). The later develop-
ment of race or ethnicity constancy may be because it is less salient than
gender in the young child’s environment (Aboud, 1988), but it may also be
due to measurement issues such as the use of perceptual transformation
measures (e.g., Aboud & Skerry, 1983), which may be more difficult than
verbal measures.
One recent study, conducted in a racially heterogeneous community, used
comparable measures for gender and racial constancy and found similar
developmental patterns for the two types of constancy (Rhee & Ruble, 1997).
Children’s understanding of gender and racial constancy developed in the
same stage progression (identity, stability, consistency). In addition, both
gender and racial constancy developed significantly with age and were correl-
ated with each other. Furthermore, when children were asked to explain their
responses, similar responses were found for the two types of constancy:
operational reasoning (e.g., “that’s the way God made him and he cannot
change”) for constant responses, and norm reasoning (e.g., “only girls have
long hair”) for nonconstant responses. Moreover, for European-American
children, understanding gender and racial constancy increased with age in a
parallel fashion. That is, in contrast to previous findings that racial constancy
developed considerably later than gender constancy, most 7-year-olds showed
constancy understanding for both social identities (83% for gender and 73%
for race). For African-American and Asian-American children, however, the
proportion of children who understood constancy was much lower for race
than for sex. For these groups, gender constancy understanding increased
with age in parallel with the European-American children, but only one in
four of the Asian-American children and one in three of the African-
American children showed racial constancy understanding by age 7.

Interpreting the pattern of developmental findings

According to social-cognitive theories, changes in children’s understanding
of social categories mirrors changes in children’s cognitive development (see,
e.g., Aboud, 1988; Kohlberg, 1969). For example, in the Rhee and Ruble
(1997) study, parallel findings in gender and racial constancy provide
supportive evidence for the role of cognition. In other words, when children
understand the Piagetian principle of conservation, such that classes of
categories, even social categories, are indicative of specific seemingly inher-
ent, immutable properties of objects and persons, children understand the
permanence of social category membership.
As argued by Katz (1983), however, children’s social cognitive processes
cannot be understood without considering both social and cognitive factors
(see also Bigler, 1995). Although cognitive factors can explain some of the
changes in social category understanding, social context factors can often
provide better explanations for some of the findings. First, gender identifica-
tion occurs prior to racial identification for both the self and others, possibly
 Ruble et al. 39
because children are constantly exposed to distinctions between boys and
girls in appearance and in behaviour from the moment they are born (Bem,
1981,1983; Katz & Kofkin, 1997).
Second, several studies report differences in majority and minority child-
ren’s social category understanding. Specifically, African-American children
have been found to be (1) more aware of race, (2) less accurate in racial
identification than European-American children, and (3) achieve racial con-
stancy later than European-American children and achieve gender constancy
prior to racial constancy (similar results were found for Asian-American
children). A social context account would seem to be implicated in these
findings, given that cognitive understanding of constancy was demonstrated
for gender. For European-American children, racial identification is simple,
as they comprise the majority group, not only in numbers but also in the
sense of representing the norm and the group in power. Within this context,
African-American children may need to be more flexible about race
(Harrison, Wilson, Pine, Chan, & Buriel, 1990; Katz & Kofkin, 1997) as they
face the challenge of identifying with both their own racial and ethnic group
and the larger European-American society (see Alejandro-Wright, 1985;
Cross, 1985).
A final, possible social-motivational explanation is that minority children’s
awareness of the social status of their respective groups leads them to either
disidentify with their group or to perceive race as changeable. Thus, consist-
ent with the verbal labelling findings described above (Katz & Kofkin, 1997),
and consistent with previous suggestions in the literature (e.g., Spencer,
1984), racial minority children may know their group membership but choose
not to view themselves as inevitably linked to their group because they sense
that it is devalued relative to the majority group. Although this hypothesis has
not been tested directly, there are some supportive findings. For example, in
one study, kindergarten and second-grade children were shown drawings of a
white, Hispanic, and black child and asked a preference question, “which one
would you like to be?” (Newman et al., 1983). The findings showed that
whereas most European-American children (especially at second grade)
selected their own group, the black and Hispanic children tended not to select
their own ethnicity, possibly suggesting that they were aware of the lower
status of their own group. To examine these alternative possible social-
motivational explanations, further future research should assess children’s
understanding of status differences more directly. In addition, future studies
(like those of Verkuyten, Chapter 7) should examine minority children
in different contexts, ones likely to vary in the extent to which they make
children feel valued, such as multicultural versus regular classrooms, or
homogeneous settings (where they are the majority) versus heterogeneous
settings (where they are in the minority). More generally, such differences
across groups raises a question about the meaning of social identities to
children of different ages and backgrounds, a topic we turn to next.
40 Children’s collective identity
Developmental changes in the meaning of a sense of “we”
What exactly does it mean when children say they are or are not a member of
a particular social group? At the earliest levels, it seems mostly to be based on
physical correlates of gender and race. In other words, children label them-
selves and others as being a member of a social category by noticing, for
example, that girls have long hair and wear dresses (Ruble & Martin, 1998).
Similarly, Spencer (1985) notes that young children use racial terms before
they understand their functional character. That is, young children become
aware of skin colour before they begin to understand that it will place them in
a particular racial group (Semaj, 1985). For instance, when young children
refer to their brown skin, it is probably not connected to being socially
labelled in our society as African-American (Spencer, 1988), nor to refer to a
racial or ethnic group of individuals (Holmes, 1995).
Once children understand constancy, however, their social category under-
standing should become more sophisticated, as they realise that changes in
physical appearance do not lead to changes in social category membership.
That is, children’s comprehension of social categories changes from one
based on superficial, physical characteristics to one based on socially con-
structed intrinsic properties. Nevertheless, even once children understand
constancy, it is still not clear to what extent such identities reflect a sense
of connection to other members of the group, including heritage, family,
country, class, and so on. Relatively little research has tried to trace the
development changes in the meaning and significance of social category
labels.
In a large study of ethnic identity in children from many different back-
grounds, Quintana (1998) characterised children’s understanding of ethnicity
and race in terms of four developmental levels. At the first level, which
emerges between the ages of 3 and 6 years, racial/ethnic differences are pri-
marily seen as physical, so that racial status is perceived as changeable (i.e.,
not constant). So, for example, if someone’s skin gets dark because they have
a tan, young children conclude that the person’s race has changed. At the
second level, which develops between the ages of 6 and 10, racial/ethnic
differences are viewed in terms of concrete cultural practices, such as food
preferences, language, and activities. To illustrate, in our ongoing study of
Russian, Dominican, and Chinese immigrant second and fourth graders,
being Chinese for one second grader meant “You just talk Chinese, not
English, and not Spanish” and being Dominican for one second grader
meant “You were born in the Dominican Republic”.
Between the ages of 10 and 14, children recognise the broader implications
of racial/ethnic differences, such as the link to social class and allocation of
resources. In other words, it is at this third level of identity that children
clearly are aware of the social status differences between racial/ethnic groups,
and the stereotypes and evaluations associated with these groups. Further-
more, children at this age often begin to identify more closely with their
 Ruble et al. 41
racial/ethnic group and want to associate more with members of this group.
Thus, this is the age when children often segregate themselves into same-race/
ethnicity friendship groups. It is not until adolescence, however, that a clearer
sense of belonging to a group is forged. Thus, it is at this fourth level of
identification that pride in racial/ethnic heritage becomes expressed.
In short, Quintana’s analysis, taken together with previous findings, sug-
gests that before the age of approximately 10 years, social categorisation and
identification is based on external characteristics, such as possessing certain
physical characteristics or engaging in particular activities. After this age, a
more sophisticated understanding of social categories develops as these cat-
egories are viewed as meaningful constructs both in society and for them-
selves. In other words, social categories are now internal social constructs
rather than simply external descriptions of what a person looks like or does.
Others do not advocate this developmental account, whereby children
under the age of 10 years simply view social categories as externally fixed
constructs. Specifically, Hirschfeld (1995) argues that young children are not
simply relying on perceptual cues when they make distinctions based on
racial categories. Whereas perceptually based theories predict that young
children’s reasoning is based on surface cues rather than abstract, non-
obvious criteria, Hirschfeld argues that young children have biologically
grounded adult-like theories of race. According to Hirschfeld, adults “appeal
to the notion that race emerges from a material but non-obvious, and
typically unspecifiable essence, part of each individual’s underlying nature
derived from group membership” (p. 217). He argues that children also have a
complex essentialised theory of race that is based on a domain-specific dis-
position to view the world in terms of discrete human kinds. In support of his
contention, Hirschfeld demonstrated that 3-year-old children expect race
(but not other characteristics such as body build) to be derived from family
background and that by 4 years of age children believe that race is fixed by
birth, biologically mediated, and immutable.
This, along with other similar findings by Hirschfeld, seems to contradict
both the racial constancy findings and Quintana’s stage model. (See, too,
Sani & Bennett, Chapter 3.) These disparate findings need not be contradict-
ory, however. It is possible that children at young ages are aware of the
biological nature of race in the sense of being sensitive to the hereditary
component of certain physical characteristics (e.g., dark skin). Yet they may
still not understand the relation between the biology of physical character-
istics of race and the immutability of the social categories of race and gender.
In other words, even though these studies ask questions that are similar on
some levels, they may in actuality be addressing different issues (i.e., biology,
stability, and meaning). Thus, future research should try to incorporate dif-
ferent measures in order to assess children’s complete understanding of racial
categories.
42 Children’s collective identity
DEVELOPMENTAL CHANGES IN THE SIGNIFICANCE OF
COLLECTIVE IDENTITY

Awareness and knowledge of one’s own and others’ category memberships is

a somewhat limited view of the development of collective identity. Social
psychologists have increasingly come to the understanding that simply know-
ing that an individual is a member of a social group is not enough to predict
the consequences that will stem from their membership. For example, accord-
ing to Tajfel and Turner’s social identity theory (1979, 1986), a person must
internalise an identity and incorporate the identity into their self-concept
before group membership will have personal consequences. Unfortunately,
models of social identity development have overlooked the need to include
such additional aspects of identifying with a social category. Highlighting
this oversight, Parke and Buriel’s (1998) critique of early ethnicity models
states, “ethnicity was equated with culture as if all members of an ethnic
group were equally involved with the ethnic culture” (p. 495). In this section,
we argue that to assess the link between collective identity and specific
developmental outcomes, collective identity, even in childhood, must be
viewed as a multidimensional construct.

Collective identity and its multidimensionality

Previous research with adolescents and adults has examined in considerable
depth the various dimensions of social identification (e.g., Luhtanen &
Crocker, 1992; Phinney, 1990) and have found that many consequences of
collective identity are mediated by the meaning a person derives from group
membership. For example, Sellers, Rowley, Chavous, Shelton, and Smith
(1997) developed the Multidimensional Model of Racial Identity (MMRI) of
African-American racial identity that reflects the multifaceted nature of
social identities. Specifically, four dimensions of racial identity are evaluated:
salience (the extent to which individuals’ race is a relevant part of their self-
concept at a particular moment in time), centrality (the extent to which indi-
viduals normatively define themselves with regard to race), ideology (the
individuals’ beliefs, opinions, and attitudes with regard to the way that the
members of their race should act), and regard (the individuals’ affective and
evaluative judgment of their race).
Each of these components of identity has unique characteristics that relate
differentially to specific outcomes. For example, Rowley, Sellers, Chavous,
and Smith (1998) found that personal regard for African-Americans pre-
dicted self-esteem only for African-American college students who reported
that race was central to their self-concept. Similarly regarding gender,
J. E. Cameron and Lalonde (2001) found that in-group ties, centrality, and
in-group affect were empirically distinct components of gender identity that
were differentially related to gender-related beliefs and behaviours. For
example, women with higher perceptions of in-group ties and greater gender
 Ruble et al. 43
centrality demonstrated more egalitarian views than women with lower
perceptions of in-group ties and less gender centrality.
We propose that developmental researchers interested in social identities
must investigate how young children think about and understand their social
group memberships in a complex and comprehensive manner, reflective of
the important findings derived from the social-psychological literature. In
the following section, we highlight some of the important social identity
dimensions (salience, centrality, knowledge, and evaluation), which we
believe clarify the link between collective identity and the overall self-concept
and subsequent developmental outcomes.

Salience
Social category salience, or the concept that individuals may be more “ready”
to perceive and process information relevant to elaborated or interconnected
categories, has been prominent in the cognitive schema literature (e.g., Bem,
1981; Martin & Halverson, 1981; Ruble & Stangor, 1986; Signorella, Bigler,
& Liben, 1993). Several different methods have been used to assess social
category salience in children. Reaction-time measures, which look at the
speed with which persons respond to social category information, have found
individual differences in the salience of gender and race (Carter & Levy, 1988;
Levy, 2000; Levy & Carter, 1989). Specifically, as early as preschool, social
category salience predicts differences in memory and stereotyping of gender
and race. In addition, researchers have used a “who said what” method,
developed by Taylor, Fiske, Etcoff, and Ruderman (1978), as an unobtrusive
measure of spontaneous categorisation. Studies using this procedure have
shown that children as young as 5 years of age spontaneously assimilate
information to social categories, and that salience effects assessed in this way
do not change between 5 and 11 years of age (Bennett & Sani, 2001; Bennett
& Sani, 2003; Bennett, Sani, Hopkins, Agostini, & Malucchi, 2000). In these
studies, salience effects were only found for social categories that have clear
perceptual cues (race and sex); for more conceptual social categories
(e.g., English vs. Scottish), no salience effect was found.
The primary means for assessing social category salience in children,
however, has been through the use of sorting tasks. When shown pictures
of people who can be classified along several social dimensions, children
prefer to sort by race and sex (e.g., Bigler & Liben, 1993; Ramsey, 1991;
Ramsey & Myers, 1990). Some studies find children using sex over race
in the sorting of photographs (e.g., Bigler & Liben, 1993), whereas other
studies find children using race over sex (e.g., Ramsey, 1991; Yee & Brown,
1988). The differential salience of these social categories is often the result
of differences in methodology (Brown, 1995). To illustrate, Davey (1983)
found that children use race when asked to sort the pictures that go
together, but use sex when asked to show which children play together. In
addition, Ramsey and Myers (1990) found that children used sex when
44 Children’s collective identity
sorting on the basis of similarity but used race when sorting on the basis of
dissimilarity.
Relative salience also depends on context and developmental level.
Previous research suggests that social identity salience varies as a function
of context, such as being the only child of a particular race or sex in a group
(e.g., McGuire, McGuire, Child, & Fujioka, 1978; Verkuyten, Chapter 7).
Similarly, when the context emphasises social categories, such as when gender
is used as a basis for classroom organisation (e.g., having boys and girls line
up separately to go out for recess), children may become predisposed to
process information in terms of gender (Bigler, 1995). Moreover, the wider
cultural or structural context may make certain social identities chronically
salient for some groups of children but not for others. For instance, Holmes
(1995) found that when children were describing their self-portraits, all of the
African-American children emphasised their skin colour, whereas European-
American children were less likely to do so. She suggests that for the African-
American children, skin colour is not simply an overt characteristic but
rather is basic to self and group identity and is highlighted by being in a
context dominated by European-American persons and culture.
In terms of the influence of development on the relative salience of gender
and race, some research suggests that sorting by race increases between 4–7
years of age, presumably in response to increasing awareness of racial
categories (Aboud, 1988; Davey, 1983). As other social information (e.g.,
race, facial expressions) becomes more informative in social judgments, sex
salience decreases (e.g., Serbin & Sprafkin, 1986; Yee & Brown, 1992). The
attainment of constancy may also influence salience. In one study, sex-related
salience was higher for children at more advanced levels of gender constancy
(Coker, 1984). In a recent study, the role of age and racial constancy in the
salience of both sex and race was assessed in 5- to 10-year-old African-
American, Asian-American, and European-American children (Rhee et al.,
2003b). When children were asked to sort photographs on the basis of simi-
larity, race was used more than sex at all age levels. When children were asked
to select the photograph that was the most like them, however, they over-
whelmingly selected on the basis of sex, but constant children selected on the
basis of race about half of the time
In short, several lines of research suggest that sex- and race-based social
categories may be salient to children at a young age. Sorting measures suggest
that race may be somewhat more salient than sex; but, as suggested by social
identity theorists, relative salience of a particular social category may depend
heavily on context (Sani & Bennett, 2001). Finally, studies examining the
development of salience show mixed findings, and the variability in age-
related trends across studies and measures suggests that more theoretically
based developmental analyses are needed. For example, some evidence sug-
gests that sex and race may become more salient as a function of increasing
knowledge about the constancy of these social categories.
How do these findings regarding the salience of social categories relate to
 Ruble et al. 45
the salience of these categories for the self-concept? Research on the devel-
opment of both the self-concept and person perception has found that social
identities are not frequently mentioned spontaneously about the self or
others (e.g., Damon & Hart, 1988; Livesley & Bromley, 1973; Ramsey, 1991).
One might predict, however, that collective identity, particularly race and
ethnicity, would be quite salient in the self-concept of some children, specific-
ally children of colour and children of recent immigrants. Research by some
of the present authors (Akiba, DiMartino, & Rodriguez, 2001a; Akiba &
Garcia Coll, in press: Akiba et al., 2001b; Alvarez, Cameron, Garfinkle,
Ruble, & Fuligni, 2001) assessed salience using measures of accessibility:
whether or not sex and racial identity was listed among the first few attributes
when children were asked to describe themselves. In these studies, children
older than 7–8 years focused heavily on personality characteristics, but not on
race/ethnicity and gender, consistent with prior research on developmental
changes in self-perceptions (e.g., Newman & Ruble, 1988). However, some
ethnic differences were found. Specifically, the Alvarez et al. study of
second- to fourth-grade Dominican and Chinese children of immigrants
found that even though most children rarely spontaneously mentioned sex,
race/ethnicity, or age, Chinese children were more likely to refer to ethnicity,
especially as they got older. Akiba and Garcia Coll found that 6- to 12-year-
old children of colour were more likely to mention race, ethnicity, or
language than were European-American children, but this difference did not
change with age, nor was it salient enough to replace the other common
sources of identity, such as activity preferences.
These data suggest that there are wide individual and group differences in
how salient various social categories are to the self-concept. Several factors
might be influential. For example, variation in homogeneity of school popu-
lations might make race or ethnicity more or less salient. In addition, self-
presentational concerns taken together with the conflictual nature of race
and ethnic identities might make it difficult for a child to mention such iden-
tities spontaneously. One might suspect that immigrant children would be
especially hesitant because of parental directives to blend in as much as pos-
sible. Thus, it will be important in future research to consider alternative
means of assessing social identity salience. In particular, measures that avoid
the sensitivity of mentioning race while at the same time allowing compar-
isons to multiple identities may be necessary to more fully understand the
developmental course of social identity salience. Furthermore, future
research must examine the moderating influence of context to assess how
individual and group differences influence collective identity salience.

Centrality
A number of investigators have identified the personal importance, strength,
or centrality of identification with a social category as a critical component
of social identity (Luhtanen & Crocker, 1992; Sellers, Smith, Shelton, Rowley,
46 Children’s collective identity
& Chavous, 1998). Centrality refers, in part, to the importance of a particular
social identity in relation to other aspects of identity, and in part, to the
frequency and preference for engaging in group-related activities and sense
of attachment to group members (Phinney, 1992). It may also be represented
in terms of a hierarchy, such that features assuming a superordinate position
are more inclusive, presumably more important, and having great implica-
tions for action (Ashmore & Ogilvie, 1992; Ogilvie & Ashmore, 1991). Cen-
trality differs from salience in that salience usually refers to a particular
moment or situation, whereas centrality refers to the extent to which indi-
viduals define themselves in terms of a particular identity across situations
(see Sellers et al., 1998). Nonetheless they are related, in that a particular
identity is likely to be salient in more situations for individuals for whom that
identity is a central part of their self-concept. In addition, if features of the
environment make a particular identity salient, it is likely to become a more
central identity in that context.
The centrality of identity has not been directly addressed in studies of
either the development of social category knowledge or self-concept. It seems
possible, however, that centrality is a key aspect of social identity constancy
and possibly the primary consequence of the attainment of constancy (see
Semaj, 1980). That is, at the point that children learn that gender or race is an
unchanging part of their self-concept, these social categories become central
aspects of their self-concept (Lutz & Ruble, 1995). Based on the develop-
mental trends in constancy described above, one might predict that gender
would become particularly central between 5–8 years of age, with race follow-
ing shortly after. An alternative developmental account, based on social
learning theory, argues that socialisation pressures to behave in accordance
with category norms is gradually internalised such that basic self-regulatory
processes maintain normative behaviour (Bussey & Bandura, 1999). Based on
this account, one might predict that a particular social identity assumes
greater centrality as self-regulatory processes assume behavioural control,
presumably at about the same age as constancy, or perhaps a bit earlier
(Bussey & Bandura, 1992). Thus, differences in socialisation could also
account for individual or group differences in centrality.
It is difficult to measure self-perceptions of centrality in young children,
and thus these hypotheses have not been directly tested. However, the studies
on children of immigrants by the present authors, mentioned earlier, have
attempted to assess social identity centrality in various ways during middle
childhood. In the Alvarez et al. (2001) study, Dominican and Chinese chil-
dren in the second–fourth grades were presented with seven social identities
on index cards in the following order: age, gender, basic ethnicity, daughter/
son, best ethnic descriptor, student, and American. Basic ethnicity referred to
the nationality of the child’s family (Dominican or Chinese). Best ethnic
descriptor was determined by presenting children with index cards labelled
with all of the ethnic words they said were about them (except American) and
then asking them to choose the word that best described them.
 Ruble et al. 47
Children were then asked three centrality measures. The first, based on
Aboud and Skerry’s (1983) measure, asked children to identify their most
important identity. Specifically, children were asked “What is the most impor-
tant thing about you, so important that without it you could no longer be
yourself ?” The results showed that best ethnic descriptor (chosen by 25% of
the children), gender (23%), and son/daughter (23%) were the most likely to
be chosen as the most important identity, and American was least likely (3%).
Interestingly, what children meant by ethnicity was generally not their basic
ethnicity (Dominican or Chinese) but rather their ethnic identity hyphenated
with American (e.g., Chinese-American, Dominican-American). Open-ended
responses suggested that children often selected these identities because it
allowed them to combine their ethnic identity with their American identity.
The second measure involved an open-ended ranking procedure, where
children were asked to rank order their social identities by placing numbers (1
through 7) on each of the index cards. The results of this ranking measure of
centrality revealed a similar pattern: son/daughter, gender, and best ethnic
descriptor were ranked highest; American was ranked the lowest. The third
measure, adapted from Willer and Romney (1988), consisted of a paired-
comparison ranking. In this procedure, children were asked a total of 15
paired-comparison questions in which they were given two social identities
and asked,“which word is more important to you?” Even in this more strin-
gent paired-comparison format children ranked the importance of their
social identities similarly. In this case, however, when the different identities
were clearly pitted against each other, son/daughter was chosen more often
relative to gender or best ethnic descriptor.
In the Akiba et al. (2001b) study, centrality was assessed using a method,
adapted from Erkut, Garcia Coll, and Alarcon (1998), in which children were
asked to indicate which of a large number of descriptors (gender, religion,
race, ethnicity, nationality) were “about them”. They were then asked to rank
order the selected labels in order of personal importance. The results showed
that European-American (41%), African-American (37%), and Asian-
American (33%) children selected boy or girl as the word that best described
them, whereas Latino-American children (26%) were more likely to select an
ethnic term (e.g., Latino). Not surprisingly, ethnicity was more “central” for
immigrant children, whereas gender was more central for American-born
children. For the next most important label, European-American children
were most likely to select “American”, whereas other children were more
likely to select race or ethnicity-related words. Thus, in this study, like the
Alvarez et al. study, both gender and race/ethnicity were viewed as quite
central relative to other possible identity labels, and race and ethnic descrip-
tors were particularly central for non-white children and for immigrant
children.
Akiba et al. (2001a) used a somewhat different measure of centrality
in their study of over 400 first- and fourth-grade children from immigrant
(Portuguese, Dominican, and Cambodian) families. Specifically, they
48 Children’s collective identity
examined the proportion of minority (e.g., Latino, Cambodian) versus
majority (e.g., white, American) labels the children selected when asked “are
you —?”. In addition, children ranked each label selected in terms of impor-
tance, as in the above two studies, and then weighted factors were created by
adding the weights for minority- versus majority-related labels. Both measures
showed an age by ethnicity interaction, as there was an age-related increase in
the centrality of minority labels and a decrease in the centrality of majority
labels only for Cambodian and Dominican children. Minority labels were
less central to Portuguese children, and relative centrality did not change
with age. In short, this study suggests that the centrality of basic ethnicity
may increase between first and fourth grades, except perhaps for white
(Portuguese) immigrant children who probably identify with American and
white with increasing age.
In summary, across several different measures and samples, these studies
indicate that gender and ethnicity are reasonably central to children during
the early to middle years of school, relative to other social identities, suggest-
ing that collective identity is a worthwhile focus of study prior to adolescence.
Furthermore, these studies demonstrate that there is consistency in children’s
selection of the social identities that are the most important to them. Not
surprisingly, however, exactly how central these collective identities were
depended on a number of other factors. First, it depended on what other
identities were available, and exactly how centrality was assessed. For example,
when the role of son or daughter was directly pitted against other social
identities in a paired-comparison format, as in the Alvarez et al. study, this
identity was clearly more central. Second, white immigrant children judged
ethnic labels as less central, possibly because outwardly they are similar to
white majority children. Finally, consistent with predictions based on develop-
mental changes in children’s understanding of racial constancy, ethnicity
but not gender centrality increased with age in certain immigrant groups.

Knowledge
Another important aspect of social identity development is children’s grow-
ing awareness of social stereotypes, status differences, and discrimination.
Virtually all children become aware of gender stereotypes, regardless of
family beliefs and values, because the messages are so blatantly presented
through the mass media and supported in peer interactions, especially at
school (Maccoby, 2002; Martin & Fabes, 2001). The developmental trend of
children’s knowledge of gender stereotypes is well documented. Children
show increasing knowledge about concrete (e.g., toys and activities) aspects
of the stereotypes until 5–6 years of age, and subsequently show an increase in
knowledge about traits associated with masculinity and femininity (Ruble &
Martin, 1998). Surprisingly little research has examined the development of
children’s knowledge of status differences, however. Children older than 10
years appear aware of greater discrimination and restrictions for women
 Ruble et al. 49
(Intons-Peterson, 1988), but the perceptions of younger children are not
known. Young children do attribute greater power to males, as Kohlberg
(1966) hypothesised, in the sense of viewing them as stronger, faster, and
more aggressive (Ruble & Martin, 1998), but it is not clear that such beliefs
reflect perceptions of higher status for males.
In contrast to the gender literature, there has been relatively little research
specifically assessing children’s racial and ethnic stereotypes, except in a
global, evaluative sense (e.g., nice–mean). Beliefs about ethnicity or race are
less likely to be as uniform as those about gender, as information is often
presented implicitly and parents vary enormously on their racial/ethnic
socialisation strategies (Garcia Coll & Pachter, in press; Hughes & Chen,
1999). Nevertheless, there is some research suggesting that young children
may have a rudimentary understanding of status differences among racial
groups (Averhart & Bigler, 1997; Katz & Kofkin, 1997). For instance, chil-
dren are aware that in America, higher status is awarded to white people and
that they are accorded more power, control, and prestige (Van Ausdale &
Feagin, 1996); even 3-year-olds may be aware of societal values regarding
race (Katz & Kofkin, 1997). Averhart and Bigler found that, even with novel
occupations, elementary-school-aged children rated those occupations
performed by blacks as lower in status than those performed by whites. In
addition, Nesdale and Flesser (2001) found that even 5-year-olds were sensi-
tive to status differences between their own and other groups and that this
difference had an impact on children’s group attitudes. Such findings have led
some to suggest that the relative status that various groups hold in society
may be reflected in children’s feelings about their own group (Spencer,
Brooklins, & Allen, 1985). For instance, “male children rarely wish to be
female, and white children never want to be black” (Katz, 1983, p. 68).
Although children do demonstrate some awareness of the status differences
of racial groups, it is still unclear exactly when and how this knowledge
affects their perceptions and evaluations of their own and other racial groups.

Evaluation
Even if two individuals label their social identity in the same way and with the
same degree of salience, centrality, and knowledge, it may be perceived in
terms of its positive qualities for one and its negative qualities for the other.
This feature of identity is associated with a number of related terms, such as
stereotypic knowledge and in-group/out-group evaluation, and it may also
be linked to intergroup prejudice. Moreover, a particularly interesting and
significant line of inquiry concerns the relation between group evaluation and
personal self-esteem (see below).
The social psychological literature has distinguished between personal
(private) regard and perceived public regard (Luhtanen & Crocker, 1992;
Sellers et al., 1998). Private regard refers to personal positive/negative feelings
about a social category (e.g., feeling that Latinos have made major
50 Children’s collective identity
accomplishments) and about being a member of that category (e.g., proud to
be a woman). Public regard refers to perceptions of others’ positive/negative
feelings (e.g., feelings that men are highly regarded by others). Although these
forms of social category evaluation might be expected to be highly related,
the degree of relation varies across groups. For example, in one study,
public and private regard were significantly associated for white and
Asian-American students but not for African-American students (Crocker,
Luhtanen, Blaine, & Broadnax, 1994). One explanation for these differences
concerns the possibility that an explicit recognition of racism may help keep
African-Americans from internalising negative messages from the broader
society (Sellers et al., 1998). Moreover, depending on the specific group,
some dimensions are more important than others in predicting well-being.
For example, in the Crocker et al. study, neither private nor public regard
predicted life satisfaction for European-American college students once
personal self-esteem was controlled; but among African-American college
students, private regard predicted life satisfaction, even with self-esteem
controlled.
A few studies on children’s perceptions of gender provide some relevant
data about developmental trends. Interestingly, most of the available evidence
suggests that young children show higher personal regard for females than for
males (Ruble & Martin, 1998). For example, Yee and Brown (1994) found
that although 3- to 11-year-olds showed the expected in-group favouritism
(see next section), both boys and girls described boys in more negative terms
overall. Ruble and Martin (1998) suggest that young children may be particu-
larly attentive to attributes that reflect moral goodness or obedience, such as
helpfulness, and that such attributes are more likely to be associated with
females. Indeed, this form of positive evaluation of females continues into
adulthood (Eagly, Mladinic, & Otto, 1991). Older children begin to view
males more highly on competency and leadership attributes (e.g., Lockheed,
Harris, & Nemceff, 1983). Thus, personal and public regard of males and
females may depend on which attributes are salient or important to the
evaluator, which probably varies as a function of age and context.
Developmental studies have typically not examined race or ethnic group
evaluation in quite this way. Most of the developmental literature that has
been interpreted in terms of evaluation refers to a comparison of in-group
versus out-group choices and preferences, rather than feelings about one’s
own group, per se. This “in-group bias” literature will be discussed below
with respect to consequences of social identity. This kind of evaluation is
potentially quite different from private regard. In-group bias can emerge even
if an identity is relatively trivial or transitory (wearing blue or yellow tee-
shirts in a classroom: Bigler et al., 1997). Thus, one might favour one’s own
group over another without experiencing a strong sense of pride in group
membership. It will be of great interest in future research to examine when
distinct and stable feelings about one’s group emerges in relation to the other
components of collective identity. It will also be interesting to examine
 Ruble et al. 51
whether the relation between private and public regard, as well as certain
differences across groups, as reported above, change in relation to emerging
knowledge about the group’s status in the broader culture.

CONSEQUENCES OF DEVELOPMENTAL CHANGES IN

COLLECTIVE IDENTITY

Social identity theory, as proposed by Tajfel and Turner (1979, 1986), high-
lights the important role that social group membership plays in shaping indi-
viduals’ self-conceptions (Turner, 1985) and interactions with both in-group
and out-group members (Tajfel, 1979). Considerable empirical research with
adults has demonstrated relations between social identity and self-evaluation,
motivation, activity engagement, and relationships with others (e.g., Crocker
et al., 1994; Sellers et al., 1998). In this section, we ask what the develop-
mental literature has to say about possible consequences of collective identity
development. Three broad types of consequences are examined: evaluative
(in-group biases and self-esteem), motivational (information-seeking and
personal choices), and interpersonal (evaluation of individual members of
groups and prejudice towards out-groups).

Evaluative consequences

In-group biases
In their original formulations of social identity theory, Tajfel and Turner
(1979) suggested that the basic process of social categorisation was sufficient
to create intergroup discrimination in favour of the in-group and against the
out-group. For example, in terms of social judgments it has been found that
people evaluate their own group members more favourably (Allen & Wilder,
1975; Oaker & Brown, 1986; Perdue, Dovidio, Gurtman, & Tyler, 1990) and
are less critical of in-group members’ behaviours (Duncan, 1976; Hewstone,
1990). Particularly when their own identity is threatened, patterns of bias
intensify and further in-group favouritism is manifested (Branscombe, Wann,
Noel, & Coleman, 1993). Behavioural ramifications include rewarding out-
group members to a lesser degree than in-group members (Brewer, 1979)
and derogating out-group members, especially when self-regard has been
threatened (Fein & Spencer, 1997).
In short, theory and research suggest that categorisation of persons into
groups becomes an evaluative process through self-identification (e.g., identi-
fication with one ethnic group and disidentification with another ethnic
group; see Hogg & Abrams, 1988, for a review). Thus, subsequent to categor-
isation, children should evaluate their own group positively and out-groups
negatively. Research on gender-related in-group bias is consistent with this
prediction (Ruble & Martin, 1998). For example, Yee and Brown (1994)
52 Children’s collective identity
found that by age 5, both boys and girls were more positive about their own
sex than about the other sex, and even 3-year-old girls showed some evidence
of in-group favouritism. Children also show greater liking for their own sex
and they typically play with same-sex others after the age of 3 (Maccoby,
1998). Finally, children assign more positive than negative traits to their own
sex during the early to middle school years (e.g., Albert & Porter, 1983;
Powlishta, 1995).
Developmental trends for race- or ethnicity-related biases differ between
children of majority versus minority status in the culture (see Aboud, 1988;
Aboud & Amato, 2001; and J. A. Cameron et al., 2001, for reviews), but these
findings provide some support for this basic hypothesis, especially for major-
ity children. White children (examined only in white-majority cultures) show
an in-group preference as early as 3–4 years of age, and in-group/out-group
differentiation increases until 7–8 years of age, either stabilising or declining
after that age (e.g., Corenblum & Annis, 1993; Doyle & Aboud, 1995). A
similar in-group bias among young children has been observed for Asian
children in an Asian society (Morland & Hwang, 1981). Children with minor-
ity status in the culture show a very different developmental pattern. Children
under 7 years of age are less likely to show an in-group preference (e.g.,
Newman et al., 1983; Spencer & Markstrom-Adams, 1990; but see Semaj,
1980, for an exception), whereas after age 7, most (but not all) studies show
in-group bias (e.g., Spencer, 1982). Thus, at about age 7, children in all groups
are showing a clear in-group positive evaluation, a pattern that is in parallel
with the emergence of racial constancy. Indeed, some research supports the
idea that the achievement of constancy understanding is associated with
greater in-group preferences, especially for ethnic/racial minority children
(Rhee et al., 2003b). What is not clear is whether such trends are related to the
different components of collective identity (e.g., centrality). Although there is
some recent evidence of clear individual differences in the degree of in-group
bias as early as preschool (Rhee, Alvarez, McGlynn, & Mull, 2003a), it is not
known whether greater in-group bias is associated with either individual
differences or developmental trends in, for example, centrality or knowledge.
Moreover, although positive evaluation of one’s group per se is obviously
part of in-group bias, it is not even clear that this evaluative component of
identity is necessarily related to in-group bias. For example, it seems possible
that one could feel proud to be a woman without feeling women are superior
to men.

Self-esteem
Social identity theory maintains that people strive for positive self-concepts
and that the emotional and value significance of social identities greatly
impacts the self-concept and self-esteem (Tajfel & Turner, 1979). By select-
ively comparing the in-group with the out-group on dimensions positive for
the in-group, one’s social identity is enhanced, thereby endowing one with a
 Ruble et al. 53
greater sense of well-being and higher self-esteem (Hogg & Abrams, 1998).
Three issues regarding this relation between collective identity and self-
esteem have dominated the literature and are relevant to a developmental
analysis.
First, the question of how individuals construct positive personal identities
that include a social identity associated with a lower status or stigmatised
group has been a major focus of research for decades. Classic “doll studies”
by Clark and Clark and others had suggested that black children identified
with the black doll but were likely to choose the white doll when asked
questions about desirable characteristics (e.g., nice, smart). These findings led
to conclusions that societal stigmatisation of a racial group is internalised as
“self-hatred” by black children, leading to lowered self-esteem. Such conclu-
sions were instrumental in the Supreme Court decision of Brown v. Board of
Education, which resulted in a social policy of racial integration.
Findings relating self-esteem to group membership have been equivocal,
however. First, a series of important conceptual and empirical analyses have
questioned the conclusion that identification with a stigmatised group leads
inevitably to lowered self-esteem (Banks, 1976; Cross, 1991) (see Phinney,
1990, for a review concerning ethnic group membership). Instead, consider-
able research has shown that individuals have at their disposal a wide range
of strategies to protect self-esteem, such as attributing negative feedback to
prejudice and devaluing dimensions on which the group fares poorly while
valuing those on which the group fares well (Crocker & Major, 1989). For
example, Steele (1988) has suggested that “self-affirmation” processes may
explain how African-American college students face stigmatisation. Because
these students do not feel that they will ever be perceived as academically
successful, they reject an identity as a hardworking student and adopt
alternative identities. Such coping strategies may maintain self-esteem, but
often at the cost of optimising the performance in valued domains of society
and perhaps lowering a sense of self-efficacy in the long run (Jones, 1992).
Nevertheless, research taking identity structure into account suggests that,
for some individuals, membership in a stigmatised group may indeed affect
self-evaluation (Rowley et al., 1998). Thus, it remains important to continue
to examine if and how identification with a devalued group lowers personal
self-esteem.
Second, despite the prevalent assumption that negative consequences stem
from attaching importance to devalued groups, group status may not be the
most important influence on the self-esteem associated with group member-
ship. Instead, because of the significance of interpersonal connections and a
sense of being accepted for self-esteem (Baumeister & Leary, 1995), identifi-
cation with a group may foster well-being, regardless of group status. For
example, some research suggests that stressful transitions may be buffered by
a strong identification with one’s own ethnic group (Ethier & Deaux, 1994).
Similar stress-buffering effects of social identity have been found in daily
diaries with adolescents (Yip & Fuligni, 2002). Collective identification may
54 Children’s collective identity
also act as a buffer for the experience of discrimination. For example, in a
study of African-American college students, Branscombe, Schmitt, and
Harvey (1999) found that the effect of making attributions of discrimination
to prejudice varied depending on collective identification. Although attribu-
tions to prejudice showed a direct relation to lower personal and collective
well-being, an indirect positive relation was also found: Attributions to
prejudice led to an increase in minority group identification that, in turn, led
to positive increases in personal and collective well-being.
Third, it follows from social identity theory that when a person’s self-
esteem is threatened, biased intergroup comparisons, intergroup discrimin-
ation, and prejudice are a likely result of self-esteem maintenance. Although
this hypothesised relation between self-esteem and prejudice is central to
social identity theory, the evidence for it has been equivocal (see Rubin &
Hewstone, 1998; cf. Fein & Spencer, 1997).
A developmental approach may be able to shed some light on these contro-
versial relations between group identification and self-esteem. To illustrate,
one would expect that the impact of group membership on self-esteem would
be problematic for children of devalued ethnic minority groups, whose ethnic
identity is central, at least once they recognise that devaluation. In one study,
for example, fifth- and sixth-grade Chinese-American children showed lower
self-esteem than first- and second-grade Chinese-American children (Ou &
McAdoo, 1999). The authors suggest that the decrease may be related to
experiences interacting with the dominant culture, making them aware of the
lower status of their racial group and affecting their sense of self-worth. It is
also possible, however, that the finding reflects other age-related differences,
because other research suggests that the self-esteem of ethnic minority
children is generally not lower than that of majority white children (e.g.,
Aboud, 1988).
We would like to suggest three reasons why it may be productive to re-
examine this relation from a developmental perspective. First, because most
studies have not examined the developmental trajectory of children’s learning
of society’s evaluations of their own and others’ ethnic groups, there is still
the possibility that they may not be capturing the key point at which there is a
drop in self-esteem prior to children’s development of coping mechanisms
(see Crocker, Major, & Steele, 1999). Second, the measures of self-esteem
employed in the past may not adequately capture the aspect influenced by
being a member of a devalued ethnic group. For example, measures that
specifically ask about wanting to change skin colour, wanting to fit in, and so
on may be more appropriate to capture self-esteem losses due to devalued
ethnic group membership (Alarcon, Szalacha, Erkut, Fields, & Garcia Coll,
2002). Moreover, it may be productive to look separately at subcomponents
of self-esteem, such as physical and social self-worth, as well as at elements of
group membership other than status, such as a sense of belonging. Finally,
racial/ethnic socialisation in the family and schools can support the develop-
ment of positive self-esteem even if the child learns about negative stereotypes
 Ruble et al. 55
applied to their group (Garcia Coll & Pachter, in press; Hughes & Chen,
1999). Future research should examine the possibility that the relation
between self-esteem and knowledge is moderated by exposure to environ-
ments that contradict negative messages.
A social identity theory perspective has typically not been applied directly
to understanding gender-related processes and self-esteem. Many studies
show that females have somewhat lower self-esteem than males (Ruble &
Martin, 1998), and sex differences in self-esteem and related constructs have
been observed prior to the well-documented sex difference in depression that
seems to emerge during adolescence (Ruble, Greulich, Pomerantz, & Goch-
berg, 1993). Other research has applied the concept of androgyny (Bem,
1974) or dualism (Spence & Helmreich, 1978) to the relation between gender
and self-esteem. The idea is that flexibility with respect to masculine and
feminine identity and behaviours is adaptive. Indeed, the literature has shown
some support for this hypothesis (see Ruble & Ruble, 1982, for a review) but,
more often, adjustment is more closely related to instrumental characteristics
or masculinity for both sexes (Aube, Norcliffe, Craig, & Koestner, 1995;
Spence & Hall, 1996). Assuming that masculinity or instrumentality are more
highly valued by society, such findings seem consistent with the idea
described above with respect to ethnicity (i.e., that negative consequences
follow from identification with a less valued group).
Recently, however, such interpretations have been questioned, and
approaches more directly relevant to social identity hypotheses have been
offered (Egan & Perry, 2001). Specifically, Egan and Perry argued that if
identification with a social group is partially in the service of self-esteem
enhancement, then feeling particularly connected or compatible with one’s
biological sex should promote feelings of well-being. In support of this
hypothesis, they found perceptions of gender typicality (e.g., “think you are a
good example of being a girl”) and gender contentedness (e.g., “like being a
girl”) were positively related to global self-worth in fourth- to eighth-graders.
Moreover, these associations remained significant when children’s percep-
tions of self-efficacy for gender-typed activities were controlled, implying that
gender identity as operationalised in this way has implications for adjustment
beyond gender-linked competencies. Interestingly, feeling pressured to engage
in gender-typed activities and in-group favouritism were negatively related to
self-esteem. This latter finding of a negative effect for in-group favouritism is
intriguing because it is inconsistent with the social identity prediction that the
value one places on one’s group should positively affect self-esteem. It will be
useful in future research to explore possible reasons for this apparent
discrepancy.

Motivational consequences
When and in what ways do children’s identification with and knowledge
about social groups influence their interests, choices, and behaviours? There
56 Children’s collective identity
has been a fair amount of research related to these questions in the gender
development literature. For example, numerous studies have examined
children’s knowledge of gender stereotypes and how that may influence their
own characteristics and behaviours (see Aubry, Ruble, & Silverman, 1999, for
a review). Little is known, however, about whether young children learn the
“do’s and don’t’s” of ethnicity in the same way. A further limitation is that
consequences of identity have been examined only with fairly limited and
unsophisticated measures of identity: simple knowledge of group distinc-
tions, or one’s placement in a group and knowledge about the stability or
constancy of those placements. The implications of the other elements of
identity, such as centrality or evaluation, have typically not been included. In
this section, we examine this literature with respect to two classes of motiv-
ational consequences: (1) interest in and information-seeking about group
norms; and (2) adherence to group norms in personal choices and behaviours.

Information-seeking and processing

Both cognitive theories regarding the development of social identity (Kohl-
berg, 1966; Martin et al., in press; Ruble, 1994) and social identity theory
(Tajfel, 1978) predict that identification with a social category has a number
of immediate consequences for children’s orientation towards other members
of the group. In particular, it should lead them to seek to be similar to other
category members and to exaggerate differences from members of contrast-
ing categories. Considerable evidence suggests that children actively seek out
and construct their own rules about gender soon after they have learnt about
gender categories and their placement in one of them (Martin et al., 2002).
For example, once children have a stable conception of gender categories,
they show increased interest in watching same-sex others (Luecke-Aleksa,
Anderson, Collins, & Schmitt, 1995; Slaby & Frey, 1975, boys only). More-
over, young children pay increased attention to and remember more informa-
tion about an activity when they believe it is something for their own sex than
for the other sex (Bradbard, Martin, Endsley, & Halverson, 1986). Finally,
there is evidence that early levels of gender identity are associated with a shift
to same-sex play (Fagot, 1985; Smetana & Letourneau, 1984).
Some research has also suggested that once racial identity becomes solidi-
fied (i.e., attainment of constancy), children engage in information-seeking
about the in-group (Aboud, 1977). It is only the previously mentioned study
by Rhee et al. (2003b), however, that directly examined differences in
information-seeking at different stages of racial constancy understanding. As
expected, the results showed that when given a choice of books to read,
children who understood the stability of their racial category were more
interested in learning about in-group characters than children who did
not have this level of understanding. In addition, African-American and
Asian-American (but not European-American) children who understood the
stability of their racial categorisation perceived themselves as more similar to
 Ruble et al. 57
other in-group members and liked them more than children who had
not reached this level of understanding. This result implies that for racial
minority children, the understanding that their racial categorisation is stable
changed the meaning of their social category membership.
In short, studies on the initial development of real group self-
categorisation are consistent with findings that manipulate assignment to
arbitrary and artificial groups. Children who have some initial understanding
of the permanence of their membership in racial and gender groups show a
more positive orientation towards group members and seek to learn more
about group norms. Future research should attempt to look at such con-
sequences longitudinally and assess why such effects often occur for some
groups only, for example, boys more than girls and racial/ethnic minority
children more than majority children.

Personal choices and behaviours

Collective identity should not only lead individuals to seek information about
group norms, but should also affect their desires to look and act in identity-
consistent ways. The motivational consequences of social group membership
take many forms and emerge from multiple pathways. For example, J. E.
Cameron (1999) has argued that social identities created by group member-
ship may guide the formation of motivationally important possible selves
(Markus & Nurius, 1986). In addition, group membership may shape per-
sonal values and interests, which in turn influences effort and performance.
For example, Eccles (1983; Wigfield & Eccles, 2000) argues that individuals
whose sex role identity is central may value those tasks associated with their
sex and devalue those tasks associated with the other sex. Similarly, Steele
(1992, 1997) suggests that identifying with a group that is stereotyped for
failing in a certain domain (e.g., academics) may lead an individual to dis-
identify in that domain in order to protect self-esteem. Moreover, experi-
mental work demonstrates that simply labelling an ambiguous activity as
being performed by one sex or the other can affect performance (see Martin
& Dinella, 2002, for a review). For example, in one study, both boys and girls
performed better on a novel throwing game when they thought is was gender
appropriate rather than gender inappropriate (Montemayor, 1974).
A developmental analysis of the processes linking personal choices and
behaviours to the emergence of collective identity seems important for several
reasons. First, identification with and knowledge about social groups emerge
in tandem during early childhood, and thus represent a unique opportunity
to examine the interplay of these two important processes. Second, the initial
emergence of social category connections and beliefs, and the context in
which they emerge, seem likely to set fundamental individual differences in
motion. Many life-defining choices – such as willingness to pursue careers
in science and math, engagement in school, orientation towards college – may
have their origins at this time. To illustrate, Steele’s research described above
58 Children’s collective identity
suggests reasons why African-Americans may perform less well in college.
But these individuals have, at least, made it to college; many more do not.
What about the children who, in essence, drop out of school as soon as they
can because they perceive school success is not for people like them? Simi-
larly, girls’ apparent loss of interest in science and maths during adolescence
may have already been predetermined by an identification with feminine
attributes and expectations at a much earlier time.
Several studies provide evidence regarding such developmental links for
gender categories. One set of studies has examined whether knowledge of
gender stereotypes affects children’s own activity preferences (see Martin et
al., in press, and Ruble & Martin, 1998, for reviews). This research has yielded
very mixed findings, indicating the inherent difficulty in examining the link
between social identity and behaviour (see Aubry et al., 1999). For example, a
large number of factors other than gender identity influence a person’s activ-
ity choice in a particular situation, such as activity attractiveness, social desir-
ability, personal skills, and familiarity. In addition, other aspects of identity,
such as centrality and evaluation, are likely to influence the extent to which
social identity guides one’s choices. Finally, developmental level has not been
carefully considered in most studies. Finding associations between knowledge
of gender stereotypes and gender-typed behavioural preferences may be pre-
cluded by the very high levels of knowledge that most children exhibit about
gender roles after the age of 5 years. It is thus noteworthy that several studies
have found significant relations between stereotyped knowledge and
behaviour/preferences (Coker, 1984; Martin, Fabes, Evans & Wyman, 1999;
Serbin, Powlishta, & Gulko, 1993). Moreover, one longitudinal study found
that many relations between knowledge and preferences involved lagged
effects (i.e., knowledge predicting preferences 1 year later; Aubry et al., 1999).
A second set of studies has examined the impact of gender identity, as
measured by children’s recognition that they are girls or boys, or higher levels
of gender identity understanding, namely stability and constancy. Consistent
with social identity theory and cognitive theories of gender development,
several studies suggest that level of understanding about one’s gender cat-
egory membership influences adherence to gender norms (see Lutz & Ruble,
1995; Martin et al., 2002, and Ruble & Martin, 1998, for reviews). There are
many inconsistencies in the findings, however, and the exact conclusion to be
drawn remains controversial (e.g., Bussey & Bandura, 1999). One promising
direction for future research would be to examine identification more directly.
For example, one could predict, in accordance with Kohlberg (1966), that
understanding the permanence of social identity would make that identity
more important or central. Another possible prediction is that the predicted
links between constancy and adherence to norms would occur only for
children who did in fact view their identity as central. It would also be of
interest in future research to look jointly at the impact of identification and
knowledge about the characteristics associated with the group. For example,
one study found that girls who had attained gender constancy and who had
 Ruble et al. 59
considerable gender stereotypic knowledge showed less interest in computers
than did all other groups of children, suggesting, perhaps that they had
abandoned an attractive activity to comply with gender norms (Newman,
Cooper, & Ruble, 1995).
There is little comparable research on the development of adherence to
ethnic norms, perhaps because activities for children are not as clearly
marked by ethnicity and race as they are for gender. One study, however,
examined differences in children’s friend and activity choices as a function of
level of racial constancy understanding (Rhee et al., 2003b). As predicted, the
results showed that children who had attained constancy were more likely to
choose a same-race activity and same-race playmates. Similar to the findings
reported above for information-seeking, these effects were stronger for the
African-American and Asian-American children than for the European-
American children.
In addition, Akiba et al. (2001a) examined the relations between social
identity and engagement in school. They found that among the Dominican-
American and Cambodian-American children, those whose “minority” iden-
tity (e.g., Dominican, Latino, Cambodian, Asian) was central tended to have
lower school engagement scores, as measured by items assessing importance
of getting good grades and importance of being liked by the teachers. These
results appear consistent with notions of stereotype threat disidentification
(Steele, Spencer, & Aronson, in press), as well as with ideas about an
oppositional identity (Cross, 1995; Fordham, & Ogbu, 1986). Other findings,
however, suggest that the proper interpretation of these relations may be
more complex and involve culture-specific meanings of engagement (Akiba
& Garcia Coll, in press). For example, Cambodian-American children with
a strong “minority” identity scored higher on the conduct aspect (e.g.,
attendance) of school.
It will be useful in future research to examine more directly the implica-
tions of children’s growing awareness of ethnicity-related ability stereotypes
for their engagement, as well as the processes underlying such relations.
Although stereotypes about the interests and characteristics of boys and girls
are well known by the age of 8 years (Ruble & Martin, 1998), there is surpris-
ingly little information about school-related ability stereotypes for either sex
or ethnicity. Nevertheless, previous research on the development of children’s
understanding of ability (e.g., Dweck, 1999) and social comparison of
abilities (Ruble & Frey, 1991) suggest that stereotype threat could appear
during middle elementary school, as long as the relevant stereotypes were
known. Indeed, recent research reported by Aronson and Good (in press) is
consistent with this prediction.

Interpersonal consequences
Considerable effort in social psychology has been devoted to showing how
identification with a particular social category (that involves a comparison
60 Children’s collective identity
with other social categories) may promote a sense of belonging and con-
nectedness but can also lead to stereotyping of out-group members and to
prejudice and intergroup conflict (Hewstone et al., 2002; Tajfel & Turner,
1979). Research in this area has been largely concerned with documenting
when and how such effects occur (e.g., whether they are automatically elicited
or can be controlled: Devine, 1989). Such research typically either examines
differences among individuals who already have relatively well-developed
social category representations or group identifications, or it varies the situ-
ation in some way, such as by increasing the salience of group distinctions.
Curiously, despite the wide-ranging importance of the interpersonal con-
sequences of social category beliefs, we know little about how these processes
develop. In this section, we consider a few areas that seem of particular
interest for a developmental analysis.

Evaluation of individual members of groups

Research with adults suggests that not all in-group members will necessarily
be evaluated positively. Specifically, individuals who deviate from in-group
social category norms are likely to be evaluated more negatively than those
who support the norms (Marques, Abrams, Paez, & Martinez-Taboada,
1998). The logic underlying this research is based upon the self-categorisation
theory idea that in-group norms represent what category members should do
and be like, and that behaving in accordance with norms helps maintain the
boundaries with other categories (Turner, Hogg, Oakes, Reicher, & Wetherell,
1987). Moreover, the group norms are legitimised by people who act in
accordance with them and such people thereby facilitate what is, for social
identity theory, a central goal of group identification: maintaining a positive
social identity (Tajfel & Turner, 1986). An interesting corollary hypothesis is
that not all out-group members will be evaluated negatively. Indeed, an out-
group deviant may be liked more than an in-group deviant, even if their
expression of norms is identical (Abrams, Marques, Bown, & Henson, 2000).
Although it is now well-documented that children show in-group biases
from an early age, it is not clear when they might adopt this more discerning
approach to interpersonal evaluation. Is there a point in development,
for example, when children’s evaluations depend on individual fit to group
norms rather than just in-group/out-group membership? Some research on
developmental changes in the inferences children draw from gender-related
information provides a partial answer to this question. In this research, chil-
dren are provided with categorical information (being male) and attribute
(typicality) information (masculine or feminine interests). In two studies,
children younger than third grade were influenced by categorical information
only, whereas older children used both types of information (Biernat, 1991;
Martin, 1989). These data suggest that children’s judgments about indi-
viduals in a group may not be influenced by information about relative typic-
ality or deviance until approximately 8 years of age or later. Future research
 Ruble et al. 61
should examine whether such developmental changes apply to perceptions of
other groups and whether they are linked in systematic ways to the growth of
social category understanding and the different dimensions of collective
identity.

Out-group derogation and prejudice

As described earlier, decades of research have shown an in-group bias for
both sex and race or ethnicity. On the basis of these results, it has been argued
that children as young as 3 years old are prejudiced (see Aboud, 1988; Brand
et al., 1974; Brown, 1995; Davey, 1983, for reviews). Indeed, in her review,
Aboud suggested that, contrary to the recent reductions seen in adult preju-
dice, the prejudice seen in children has remained constant.
We have argued elsewhere (J. A. Cameron et al., 2001), however, that it is
premature to conclude from this research that young children are exhibiting
prejudice. Developmental researchers, in confounding attitudes towards in-
groups and out-groups, have not incorporated recent propositions that posi-
tive in-group bias is not the same as prejudice (Brewer, 1999, 2001). In our
review of the literature on prejudice in children, we found that much of the
past research in this area relied on measures and/or statistical analyses that
confounded in-group positivity and out-group negativity (J. A. Cameron et
al., 2001). Thus, in these studies it was not possible to determine whether
children’s differential responses were the result of out-group derogation
rather than simple in-group favouritism. The few studies that had employed
measures that separately assessed evaluations towards the in-group and out-
group demonstrated that evidence of out-group derogation was seen only in
older children (12-year-olds). Therefore, it seems likely that young children,
rather than expressing prejudice, are probably manifesting in-group
favouritism.
The lack of an empirical relationship between in-group favouritism and
out-group derogation suggests that these two aspects of intergroup behaviour
may have different motivational origins (Brewer, 1999; Brown, 1995). Brewer
(2001) suggests a hierarchy such that the principles of social categorisation
and in-group positivity are most likely to be universal aspects of human
social groups. Intergroup comparison and out-group hostility, in contrast,
“require additional social-structural and motivational conditions that are not
inherent in the processes of group formation itself” (Brewer, 2001, p. 19). For
example, prejudice against women is more likely when they violate gender
stereotypes or participate in male-dominated domains (Eagly & Mladinic,
1994). As Lutz and Ruble (1995) illustrate: “girls’ efforts to ‘beat the boys’ in
areas where boys may be tacitly expected to do better, such as playground
athletics, are likely to provoke resentment and active attempts at exclusion on
the part of boys” (p. 144).
We further suggest that future research should examine whether the
developmental trajectory of intergroup attitudes may be described similarly
62 Children’s collective identity
to Brewer’s (2001) hierarchical progression of relations between in-group
formation and intergroup behaviour. That is, when children understand
social categorisation, they exhibit in-group positivity. Only with sufficient
social-cognitive development and specific social-structural and motivational
conditions, however, will children exhibit intergroup comparisons and out-
group hostility. According to this framework, prejudice would emerge as an
interaction of developmental changes and social-contextual factors, such as
the explicit socialisation of evaluations associated with particular groups or
more implicit socialisation that membership in particular groups has a
meaningful function.
In some cases, explicit socialisation practices may lead even young children
to develop out-group derogation. Such messages might be conveyed to young
children by racist parents or in mass media portrayals of conflict with other
nationalities or ethnicities. Indeed, when examining ethnic attitudes in a
country where clear ethnic friction results in explicit out-group derogation
(e.g., the Arab–Israeli conflict), it becomes evident that children can be social-
ised to view certain groups in a prejudicial manner at a very young age. For
instance, Bar-Tal (1996) found that Israeli children as young as 2½ years old
rated a male photo more negatively when he was identified as an Arab than
when the photo was not labelled. Similarly, Rutland (1999) found evidence of
out-group derogation for children’s evaluations of a group that has been
depicted negatively in British society (Germans) (see also Bennett, Lyons,
Sani, & Barrett, 1998). Thus, it is likely that young children exposed to fre-
quent and overt negativity towards particular groups may develop out-group
negativity as well as in-group positivity.
In addition, prejudicial attitudes could result from children learning that
social categories have functional value (Bigler, 1995; Bigler et al., 1997). For
instance, Bigler (1995) found that children exhibited more occupational
stereotyping of men and women in a classroom when gender was made func-
tional by the teacher. In more recent research, novel groups, unequal in status,
were created and made salient by teachers. Children who were members of
high-status groups developed in-group biased attitudes (Bigler, Spears-
Brown, & Markell, 2001). These studies suggest that social conditions, such
as the differences in social status and the extent to which group membership
influences children’s experiences, affect the extent to which children make
competitive intergroup comparisons. Hence, when society makes functional
use of racial/ethnic groups, prejudice is a conceivable consequence; but
whether or not such conditions directly foster out-group hostility is a
question that awaits future research.
In addition to context, researchers need to account for the role of social
category understanding and multidimensionality of collective identity in
determining when children might demonstrate out-group derogation (J. A.
Cameron et al., 2001). Even though young children (under the age of 7 years)
might demonstrate some out-group devaluation, these attitudes will be more
rudimentary and unlike the prejudicial attitudes of older children and adults
 Ruble et al. 63
(see also Rhee et al., 2003a). That is, aspects that would underlie prejudice –
that individuals belong to distinct, immutable racial groups with particular
interests and proclivities that endure over time – are not likely to be available
to young children until they have realised certain social-cognitive capacities
such as racial constancy, an understanding of stable traits, and the ability to
engage in social comparisons. Moreover, it will be of interest in future
research to examine the interplay between context and the emergence of the
different dimensions of collective identity as it influences the development of
prejudicial attitudes.

SUMMARY AND CONCLUSIONS

Given the salience and consequences of group membership and social identi-
fication during adolescence and young adulthood, the lack of attention to the
development of collective identity during childhood is surprising and
unfortunate. Our review of related developmental literatures suggests that
identification with social groups, especially gender and ethnicity or race, is
likely to be quite significant to children by middle childhood. First, develop-
mental trends observed in research on children’s development of a personal
self implies that between 5 and 9 years of age, social categories – particularly
gender and race – should change from being viewed in terms of overt physical
characteristics (e.g., long hair, dark skin) to becoming a comparison of
traits and behaviours and, in the process, take on evaluative meaning. Thus,
collective identities may become quite significant to children’s self-concept by
middle childhood. Second, research on children’s understanding of and iden-
tification with social categories suggests that children are capable of a sense
of “we” by preschool but that the nature and meaning of this identification
changes dramatically during childhood and varies as a function of ethnic/
racial minority or majority status. Third, borrowing from social-psycho-
logical research and theory, we further argued that simply labelling oneself as
a member of a particular group is inadequate to fully capture the significance
of collective identity. Instead, we adopted a multidimensional view in which
salience, centrality, knowledge, and evaluation were viewed as integral but
separable components of this identity. Although there is relatively little
developmental work to date on the emergence of these components of col-
lective identity, recent research suggests that gender and ethnicity are reason-
ably central to children during the early to middle years of school, relative to
other social identities, and that there is consistency in children’s selection of
the social identities that are the most important to them. It also appears that
the relative importance of ethnicity increases with age in some groups. These
conclusions about centrality are preliminary, however, and, with the excep-
tion of clear evidence regarding the development of gender stereotypes by
middle childhood, we do not yet know much about the development of
the other components of collective identity. Given the clear personal and
64 Children’s collective identity
interpersonal consequences later in development, we view focusing on these
different components of collective identity as a high priority for future
research.

Acknowledgements
We appreciate the helpful comments and suggestions made on an earlier draft
by Anna Akerman and Tracy McLauglin-Volpe. We are also grateful to Faith
Greulich for providing support in manuscript preparation. Preparation of the
chapter was facilitated by grants from the National Institute of Mental
Health (MH37215), MacArthur Foundation, and Russell Sage Foundation.

REFERENCES

Aboud, F. E. (1977). Interest in ethnic information: A cross-cultural developmental

study. Canadian Journal of Behavioural Science, 9, 134–146.
Aboud, F. E. (1984). Social and cognitive bases of ethnic identity. Journal of Genetic
Psychology, 145, 217–230.
Aboud, F. E. (1988). Children and prejudice. Oxford: Basil Blackwell.
Aboud, F. E., & Amato, M. (2001). Developmental and socialization influences on
intergroup bias. In R. Brown & S. Gaertner (Eds.), Blackwell handbook in social
psychology: Intergroup processes, Vol. 4. Oxford: Blackwell.
Aboud, F., & Ruble, D. N. (1987). Identity constancy in children: Developmental
processes and implications. In T. M. Honess & K. M. Yardley (Eds.), Self and
identity: Individual change in development (pp. 95–107). London: Routledge &
Kegan Paul.
Aboud, F. E., & Skerry, S. A. (1983). Self and ethnic concepts in relation to ethnic
constancy. Canadian Journal of Behavioral Science, 15, 14–26.
Abrams, D., Marques, J. M., Bown, N., & Henson, M. (2000). Pro-norm and
anti-norm deviance within and between groups. Journal of Personality and Social
Psychology, 78, 906–912.
Akiba, D., DiMartino, L., & Rodriquez, S. (2001a). The construction of self among
children of color and children from immigrant families. Paper presented at the
biennial meeting of the Society for Research in Child Development in Minneapolis,
MN.
Akiba, D., & Garcia Coll, C. (in press). Effective interventions with children of color
and their families: An ecological systems approach. In T. B. Smith (Ed.), Handbook
of multicultural counseling: Internalizing and affirming diversity in counseling and
psychology. Boston: Allyn & Bacon.
Akiba, D., Garcia Coll, C., & Magnuson, K. (2001b). Children of color and children
from immigrant families: The development of social identities, school engagement and
interethnic social attribution during middle childhood. Manuscript under review.
Alarcon, O., Szalacha, L. A., Erkut, S., Fields, J. P., & Garcia Coll, C. T. (2002). The
color of my skin: A measure to assess children’s perceptions of their skin color.
Applied Developmental Science, 14, 208–221.
Albert, A. A., & Porter, J. R. (1983). Age patterns in the development of children’s
gender-role stereotypes. Sex Roles, 9, 59–67.
 Ruble et al. 65
Alejandro-Wright, M. N. (1985). The child’s conception of racial classification: A
socio-cognitive developmental model. In M. D. Spencer, G. K. Brookins, & W. R.
Allen (Eds.), Beginnings: The social and affective development of black children
(pp. 185–200). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Allen, V. L., & Wilder, D. A. (1975). Categorization, belief similarity, and intergroup
discrimination. Journal of Personality and Social Psychology, 32, 971–977.
Alvarez, J. M., Cameron, J., Garfinkle, G. S., Ruble, D. N., & Fuligni, A. J. (2001, April).
Identity development in immigrant children. Paper presented at the biennial meeting
of the Society for Research in Child Development. Minneapolis, MN.
Alvarez, J. M., Ruble, D. N., & Bolger, N. (2001). The role of evaluation in the
development of person perception, Child Development, 72, 1409–1425.
Aronson, J., & Good, C. (in press). The development and consequences of stereotype
vulnerability in adolescents. In F. Pajaras & T. Urdan (Eds.), Adolescence and
education, Vol. 2: Academic motivation of adolescents. Greenwich, CT: Information
Age Publishing.
Ashmore, R. D., Jussim, L., Wilder, D., & Heppen, J. (2001). Toward a social identity
framework for intergroup conflict. In R.D. Ashmore, L. Jussim, & D. Wilder
(Eds.), Social identity, intergroup conflict, and conflict reduction. London: Oxford
University Press.
Ashmore, R. D., & Ogilvie, D. M. (1992). He’s such a nice boy . . . when he’s with
grandma: Gender and evaluation in self-with-other representations. In T. M.
Brinthaupt & R. P. Lipka (Eds.), The self: Definitional and methodological issues
(pp. 236–290). Albany, NY: State University of New York Press.
Aube, J., Norcliffe, H., Craig, J., & Koestner, R. (1995). Gender characteristics and
adjustment-related outcomes: Questioning the masculinity model. Personality and
Social Psychology Bulletin, 21, 284–295.
Aubry, S., Ruble, D. N., & Silverman, L. (1999). The role of gender knowledge in
children’s gender-typed preferences. In C. Tamis-LeMonda & L. Balter (Eds.),
Child psychology: A handbook of contemporary issues. Philadelphia, PA: Psychology
Press.
Averhart, C., & Bigler, R. (1997). Shades of meaning: Skin tone, racial attitudes, and
constructive memory in African American children. Journal of Experimental Child
Psychology, 67, 363–388.
Banks, W. C. (1976). White preference in blacks: A paradigm in search of a
phenomenon. Psychological Bulletin, 83, 1179–1186.
Bar-Tal, D. (1996). Development of social categories and stereotyping in early
childhood: The case of “the Arab” concept formation, stereotype and attitudes by
Jewish children in Israel. International Journal of Intercultural Relations, 20,
341–370.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal
attachments as a fundamental human motivation. Psychological Bulletin, 117,
497–529.
Bem, S. L. (1974). The measurement of psychological androgyny. Journal of
Consulting and Clinical Psychology, 42, 155–162.
Bem, S. L. (1981). Gender schema theory: A cognitive account of sex typing.
Psychological Review, 88, 354–364.
Bem, S. L. (1983). Gender schema theory and its implication for child development:
Raising gender-aschematic children in a gender-schematic society. Signs, 8,
598–616.
66 Children’s collective identity
Bennett, M., Lyons, E., Sani, F., & Barrett, M. (1998). Children’s subjective
identification with the group and in-group favouritism. Developmental Psychology,
34, 902–909.
Bennett, M., & Sani, F. (2001). Children’s social categorization: Evidence for automatic
processing of ethnicity. Paper presented at the biennial meeting of the SRCD in
Minneapolis, MN.
Bennett, M., & Sani, F. (2003). The role of target gender and race in children’s
encoding of category-neutral person information. British Journal of Developmental
Psychology, 21, 99–112.
Bennett, M., Sani, F., Hopkins, N., Agostini, L., & Malucchi, L. (2000). Children’s
gender categorization: An investigation of automatic processing. British Journal of
Developmental Psychology, 18, 97–102.
Bernal, M. E., & Knight, G. P. (Eds.) (1993). Ethnic identity: Formation and trans-
mission among Hispanics and other minorities. Albany, NY: State University of New
York Press.
Bernal, M. E., & Knight, G. P. (1997). Ethnic identity of Latino children. In J. G.
Garcia & M. C. Zea (Eds), Psychological interventions and research with Latino
populations (pp. 15–38). Needham Heights, MA: Allyn & Bacon.
Bernal, M. E., Knight, G. P., Garza, C. A., Ocampo, K. A., & Cota, M. K. (1990). The
development of ethnic identity in Mexican-American children. Hispanic Journal of
Behavioral Sciences, 12, 3–24.
Biernat, M. (1991). Gender stereotypes and the relationship between masculinity and
femininity: A developmental analysis. Journal of Personality and Social Psychology,
61, 351–365.
Bigler, R. S. (1995). The role of classification skill in moderating environmental
influences on children’s gender stereotyping: A study of the functional use of
gender in the classroom. Child Development, 66, 1072–1087.
Bigler, R. S. (1997). Conceptual and methodological issues in the measurement of
children’s sex typing. Psychology of Women Quarterly, 21, 53–69.
Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorization and
the formation of intergroup attitudes in children. Child Development, 68,
530–543.
Bigler, R. S., & Liben, L. S. (1993). Cognitive mechanisms in children’s gender
stereotyping: Theoretical and educational implications of a cognitive-based
intervention. Child Development, 63, 1351–1363.
Bigler, R. S., Spears-Brown, C., & Markell, M. (2001). When groups are not created
equal: Effects of group status on the formation of intergroup attitudes in children.
Child Development, 72, 1151–1162.
Bowlby, J. (1980). Attachment and loss: Vol. 1. Attachment (original work published in
1969). New York: Basic Books.
Bradbard, M., Martin, C., Endsley, R., & Halverson, C. (1986). Influence of
sex stereotypes on children’s exploration and memory: A competence versus
performance distinction. Developmental Psychology, 22, 481–486.
Brand, E. S., Ruiz, R. A., & Padilla, A. M. (1974). Ethnic identification and
preference: A review. Psychological Bulletin, 81, 860–890.
Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999). Perceiving pervasive
discrimination among African Americans: Implications for group identification
and well-being. Journal of Personality and Social Psychology, 77, 135–149.
Branscombe, N. R., Wann, D. L., Noel, J. G., & Coleman, J. (1993). In-group or
 Ruble et al. 67
out-group extremity: Importance of the threatened social identity. Personality and
Social Psychology Bulletin, 19, 381–388.
Bretherton, I., McNew, S., & Beeghly-Smith, J. (1981). Early person knowledge as
expressed in gestural and verbal communication: When do infants acquire a
“theory of mind”? In M. E. Lamb & L. R. Sherrod (Eds.), Infants’ social cognition
(pp. 333–373). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Brewer, M. B. (1979). Ingroup bias in the minimal intergroup situation: A cognitive-
motivational analysis. Psychological Bulletin, 86, 307–324.
Brewer, M. B. (1991). The social self: On being the same and different at the same
time. Personality and Social Psychology Bulletin, 17, 475–482.
Brewer, M. B. (1999). The psychology of prejudice: Ingroup love or outgroup hate?
Journal of Social Issues, 55: 429–444. (Special Issue: Prejudice and intergroup
relations: Papers in honour of Gordon W. Allport’s centennial.)
Brewer, M. B. (2001). Ingroup identification and intergroup conflict: When does
ingroup love become out-group hate? In R. D. Ashmore, L. Jussim, & D. Wilder
(Eds.), Social indentity, intergroup conflict, and conflict reduction. New York: Oxford
University Press.
Brewer, M. B., & Brown, R. J. (1998). Intergroup relations. In D. T. Gilbert, S. T. Fiske
et al. (Eds.), The handbook of social psychology, Vol. 2 (4th ed.; pp. 554–594). New
York: McGraw-Hill.
Brewer, M. B., & Gardner, W. (1996). Who is this we? Levels of collective identity and
self representations. Journal of Personality and Social Psychology, 71, 83–93.
Brown, R. (1995). Prejudice: Its social psychology. Oxford: Blackwell.
Bussey, K., & Bandura, A. (1992). Self-regulatory mechanisms governing gender
development. Child Development, 63, 1236–1250.
Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and
differentiation. Psychological Review, 106, 676–713.
Cameron, J. A., Alvarez, J. M., Ruble, D. N., & Fuligni, A. J. (2001). Children’s
lay theories about ingroups and outgroups: Reconceptualizing research on
“prejudice”. Personality and Social Psychology Review, 5, 118–128.
Cameron, J. E. (1999). Social identity and the pursuit of possible selves: Implications
for the psychological well-being of university students. Group Dynamics, 3,
179–189.
Cameron, J. E. & Lalonde, R.N. (2001). Social identification and gender related ideol-
ogy in men and women. British Journal of Social Psychology, 40, 59–77.
Carter, D. B., & Levy, G. D. (1988). Cognitive aspects of children’s early sex-role
development: The influence of gender schemas on preschoolers’ memories and
preferences for sex-typed toys and activities. Child Development, 59, 782–793.
Coker, D. R. (1984). The relationships among gender concepts and cognitive maturity
in preschool children. Sex Roles, 10, 19–31.
Connolly, P. (1998). Racism, gender identities and young children: social relations in a
multi-ethnic, inner-city primary school. New York: Routledge.
Cooley, C. H. (1902). Human nature and the social order. New York: Scribner.
Corenblum, B., & Annis, R. C. (1993). Development of racial identity in minority and
majority children: An affect discrepancy model. Canadian Journal of Behavioural
Science, 25, 499–521.
Crocker, J., Luhtanen, R., Blaine, B., & Broadnax, S. (1994). Collective self-esteem
and psychological well-being among white, black, and Asian college students.
Personality and Social Psychology Bulletin, 20, 503–513.
68 Children’s collective identity
Crocker, J., & Major, B. (1989). Social stigma and self-esteem: The self-protective
properties of stigma. Psychological Review, 96, 608–630.
Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. T. Gilbert, S. T. Fiske,
& G. Lindzey (Eds.), Handbook of social psychology (4th ed.; pp. 504–553). New
York: McGraw-Hill.
Cross, W. E. (1978). The Thomas and Cross models of psychological nigrescence: A
literature review. Journal of Black Psychology, 5, 13–31.
Cross, W. E. (1985). Black identity: Rediscovering the distinction between personal
identity and reference group orientation. In M. B. Spencer, G. K. Brookins, & W. R.
Allen (Eds.), Beginnings: The social and affective development of black children
(pp. 155–171). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Cross, W. E. (1991). Shades of black: Diversity in African American identity. Madison,
WI: University of Wisconsin Press.
Cross, W. E. (1995). Oppositional identity and African American youth: Issues and
prospects. In W. Hawley & A. Jackson (Eds.), Toward a common destiny: Improving
race and ethnic relations in America (pp. 185–204). San Francisco: Jossey-Bass.
Cross, W. E. Jr., & Fhagen-Smith, P. (2001). Patterns of African American identity
development: A life span perspective. In B. Jackson & C. Wijeyesinghe (Eds.), New
perspectives on racial identity development. New York: New York University Press.
Cross, W. E, Strauss, L., & Fhagen-Smith, P. (1999). African American identity
development across the life span: Educational implications. In R. Hernandez &
E. R. Hollins (Eds.), Racial and ethnic identity in school practices: Aspects of human
development (pp. 29–47). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Damon, W., & Hart, D. (1988). Self-understanding in childhood and adolescence.
Cambridge: Cambridge University Press.
Davey, A. (1983). Learning to be prejudiced: Growing up in multi-ethnic Britain.
London: Edward Arnold.
Deaux, K., & Martin, D. (2001). Which context? Specifying levels of context in identity
processes. Paper presented at Indiana Conference on Identity Theory.
Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled
components. Journal of Personality and Social Psychology, 56, 5–18.
Doyle, A. B., & Aboud, F. E. (1995). A longitudinal study of white children’s racial
prejudice as a social-cognitive development. Merrill-Palmer Quarterly, 41, 209–228.
Duncan, B. L. (1976). Differential social perception and attribution of intergroup
violence: Testing the lower limits of stereotyping of blacks. Journal of Personality
and Social Psychology, 34, 590–598.
Durkin, K. (1995). Developmental social psychology: From infancy to old age.
Cambridge, MA: Blackwell Publishers.
Dweck, C. S. (1999). Self-theories: Their role in motivation, personality, and
development (Essays in Social Psychology). Philadelphia, PA: Psychology Press.
Eagly, A. H., & Mladinic, A. (1994). Are people prejudiced against women? Some
answers from research on attitudes, gender stereotypes, and judgments of
competence. In W. Stroebe & M. Hewstone (Eds.), European Review of Social
Psychology: Vol. 5 (pp. 1–35). New York: John Wiley & Sons.
Eagly, A. H., Mladinic, A., & Otto, S. (1991). Are women evaluated more favourably
than men? An analysis of attitudes, beliefs, and emotions. Psychology of Women
Quarterly, 15, 203–216.
Eccles, J. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.),
Achievement and achievement motives. San Francisco, CA: W. H. Freeman.
 Ruble et al. 69
Eder, R. A. (1990). Uncovering young children’s psychological selves: Individual and
developmental differences. Child Development, 61, 849–863.
Egan, S. K., & Perry, D. G. (2001). Gender identity: A multidimensional analysis with
implications for psychosocial adjustment. Developmental Psychology, 37, 451–463.
Erikson, E. H. (1963). Childhood and society (2nd ed.). New York: Norton.
Erkut, S., Garcia Coll, C., & Alarcon, O. (1998). Ethnic self-identity. A measure to
assess children’s conceptualization of race and ethnicity. Manuscript in preparation.
Ethier, K. A., & Deaux, K. (1994). Negotiating social identity when contexts change:
Maintaining identification and responding to threat. Journal of Personality and
Social Psychology, 67, 243–251.
Fagan, J. F., & Singer, L. T. (1979) The role of simple feature differences in infants’
recognition of faces. Infant Behavior and Development, 2, 39–45.
Fagot, B. I. (1985). Changes in thinking about early sex role development.
Developmental Review, 5, 83–98.
Fein, S., & Spencer, S. (1997). Prejudice as self-image maintenance: Affirming the self
through derogating others. Journal of Personality and Social Psychology, 73, 31–44.
Fordham, C., & Ogbu, J. (1986). Black students’ school success: Coping with the
burden of “acting white”. Urban Review, 18, 176–206.
Garcia Coll, C. T., & Pachter, L. (in press). Ethnic and minority parenting. In M. H.
Bornstein (Ed.), Handbook of parenting (2nd ed.). Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Gelman, S. A., & Markman, E. M. (1986). Categories and inductions in young
children. Cognition, 23, 183–209.
Harrison, A. O., Wilson, M. N., Pine, C. J., Chan, S. Q., & Buriel, R. (1990). Family
ecologies of ethnic minority children. Child Development, 61, 347–362.
Harter, S. (1983). Developmental perspectives on the self-system. In E. M.
Hetherington (Ed.), Handbook of child psychology: Vol. 4. Socialization, personality,
and social development. New York: John Wiley & Sons.
Harter, S. (1998). The development of self-representations. In W. Damon & N.
Eisenberg (Eds.), Handbook of child psychology: Vol. 3. Social, emotional, and
personality development (pp. 553–617). New York: John Wiley & Sons.
Hewstone, M. (1990). The “ultimate attribution error”? A review of the literature on
intergroup causal attribution. European Journal of Social Psychology, 20, 311–335.
Hewstone, M., Rubin, M., & Willis, H. (2002). Intergroup bias. Annual Review of
Psychology, 53, 575–604.
Higgins, E. T. (1991). Development of self-regulatory and self-evaluative processes:
Costs, benefits, and tradeoffs. In M. R. Gunnar & L. A. Sroufe (Eds.), Self processes
and development: The Minnesota Symposium on Child Psychology: Vol. 23,
pp. 125–166. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Higgins, E. T., & Parsons, J. E. (1983). Social cognition and the social life of the child:
Stages as subcultures. In E. T. Higgins, D. N. Ruble, & W. W. Hartup (Eds.), Social
cognition and social development (pp. 15–62). New York: Cambridge University
Press.
Hirschfeld, L. A. (1995). Do children have a theory of race? Cognition, 54,
209–252.
Hogg, M. A. (2000). Social identity and social comparison. In J. Suls & L. Wheeler
(Eds), Handbook of social comparison: Theory and research. The Plenum series in
social/clinical psychology (pp. 401–421). New York: Kluwer Academic/Plenum
Publishers.
70 Children’s collective identity
Hogg, M. A., & Abrams, D. (1988). Social identifications: A social psychology of
intergroup relations and group processes. London: Routledge.
Holmes, R. M. (1995). How young children perceive race. Thousand Oaks, CA:
SAGE Publications.
Hughes, D., & Chen, L. (1999). The nature of parents’ race-related communications
to children: A developmental perspective. In L. Balter & C. S. Tamis-LeMonda
(Eds.), Child psychology: A handbook of contemporary issues (pp. 467–490).
Philadelphia, PA: Psychology Press/Taylor & Francis.
Intons-Peterson, M. J. (1988). Gender concepts of Swedish and American youth.
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Jones, J. M. (1992). Understanding the mental health consequences of race:
Contributions of basic social psychological processes. In D. N. Ruble, P. R.
Costanzo, & M. E. Oliveri (Eds.), The social psychology of mental health: Basic
mechanisms and applications (pp. 199–240). New York: Guilford Press.
Katz, P. A. (1983). Developmental foundations of gender and racial attitudes. In R. L.
Leahy (Ed.), The child’s construction of social inequality (pp. 41–78). New York:
Academic Press.
Katz, P. A., & Kofkin, J. A. (1997). Race, gender, and young children. In S. S. Luthar,
J. A. Burack, D. Cicchetti, & J. Weisz (Eds.), Developmental psychopathology:
Perspectives on adjustment, risk, and disorder (pp. 51–74). New York: Cambridge
University Press.
Kohlberg, L. A. (1966). A cognitive-developmental analysis of children’s sex role
concepts and attitudes. In E. E. Maccoby (Ed.), The development of sex differences
(pp. 82–173). Stanford, CA: Stanford University Press.
Kohlberg, L. (1969). Stage and sequence: The cognitive-developmental approach to
socialization. In D. A. Goslin (Ed.), Handbook of socialization theory and research
(pp. 347–480). Chicago, IL: Rand McNally.
Leinbach, M. D., & Fagot, B. I. (1986). Acquisition of gender labelling: A test for
toddlers. Sex Roles, 15, 655–666.
Leinbach, M. D., & Fagot, B. I. (1993). Categorical habituation to male and female
faces: Gender schematic processing in infancy. Infant Behavior and Development, 16,
317–332.
Levy, G. D. (2000). Individual differences in race schematicity as predictors of African
American and white children’s race-relevant memories and peer preferences. The
Journal of Genetic Psychology, 16, 400–419.
Levy, G. D., & Carter, D. B. (1989). Gender schema, gender constancy, and
gender-role knowledge: The roles of cognitive factors in preschoolers’ gender-role
stereotype attributions. Developmental Psychology, 25, 444–449.
Lewis, M. (1990). Social knowledge and social development. Merrill-Palmer
Quarterly, 36, 93–116.
Livesley, W., & Bromley, D. (1973). Person perception in childhood and adolescence.
London: John Wiley & Sons.
Lockheed, M. E., Harris, A. M., & Nemceff, W. P. (1983). Sex and social influence:
Does sex function as a status characteristic in mixed-sex groups of children? Journal
of Educational Psychology, 75, 877–888.
Luecke-Aleksa, D., Anderson, D. R., Collins, P. A., & Schmitt, K. L. (1995). Gender
constancy and television viewing. Developmental Psychology, 31, 773–780.
Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self-evaluation of
one’s social identity. Personality and Social Psychology Bulletin, 18, 302–318.
 Ruble et al. 71
Lutz, S. E., & Ruble, D. N. (1995). Children and gender prejudice: Context,
motivation, and the development of gender concepts. Annals of Child Development,
10, 131–166.
Maccoby, E. E. (1988). Gender as a social category. Developmental Psychology, 24,
755–765.
Maccoby, E. E. (1998). The two sexes: Growing up apart, coming together. Cambridge,
MA: Harvard University Press.
Maccoby, E. E. (2002). Gender and group process. Current Directions in Psychological
Science, 11, 54–58.
Marcia, J. (1988). Identity, cognitive/moral development, and individuation. In D. K.
Lapsley & F. C. Power (Eds.), Self, ego, and identity: Integrative approaches
(pp. 211–225). New York: Springer-Verlag.
Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist, 41, 954–969.
Marques, J., Abrams, D., Paez, D., & Martinez-Taboada, C. (1998). The role of
categorization and in-group norms in judgments of groups and their members.
Journal of Personality and Social Psychology, 75, 976–988.
Marques, J. M., Robalo, E. M., & Rocha, S. A. (1992). Ingroup bias and the “black
sheep” effect: Assessing the impact of social identification and perceived variability
on group judgments. European Journal of Social Psychology, 22, 331–352.
Martin, C. L. (1989). Children’s use of gender-related information in making social
judgments. Developmental Psychology, 25, 80–88.
Martin, C. L., & Dinella, L. M. (2001). Children’s gender cognitions, the social
environment, and sex differences in cognitive domains. In A. M. Lisi & R. D. Lisi
(Eds.), Biology, society, and behavior: The development of sex differences in cognition
(pp. 207–239). Westport, CT: Ablex.
Martin, C. L., & Dinella, L. M. (2002). Children’s gender cognitions, the social
environment, and sex differences in cognitive domains. In A. V. McGillicuddy-
De Lisi & R. De Lisi (Eds.), Biology, society, and behavior: The development of
sex differences in cognition. Advances in applied developmental psychology, Vol. 21
(pp. 207–239). Westport, CT: Ablex.
Martin, C. L., & Fabes, R. A. (2001). The stability and consequences of young child-
ren’s same-sex peer interactions. Developmental Psychology, 37, 431–446.
Martin, C. L., Fabes, R. A., Evans, S. M., & Wyman, H. (1999). Social cognition on
the playground: Children’s beliefs about playing with girls versus boys and their
relations to sex segregated play. Journal of Social and Personal Relationships, 16,
751–771.
Martin, C. L., & Halverson, C. F. (1981). A schematic processing model of sex typing
and stereotyping in children. Child Development, 52, 1119–1134.
Martin, C. L., Ruble, D. N., & Szkrybalo, J. (2002). Cognitive theories of early gender
development. Psychological Bulletin, 128, 903–933.
McGuire, W. J., McGuire, C. V., Child, P., & Fujioka, T. (1978). Salience of ethnicity
in the spontaneous self-concept as a function of one’s ethnic distinctiveness in the
social environment. Journal of Personality and Social Psychology, 36, 511–520.
Mead, G. H. (1934). Mind, self, and society: From the standpoint of social behaviorist.
Chicago, IL: University of Chicago Press.
Montemayor, R. (1974). Children’s performance in a game and their attraction to it as
a function of sex-typed labels. Child Development, 45, 152–156.
Morland, J. K., & Hwang, C. H. (1981). Racial/ethnic identity of preschool children.
Journal of Cross-Cultural Psychology, 12, 409–424.
72 Children’s collective identity
Nesdale, D., & Flesser, D. (2001). Social identity and the development of children’s
group attitudes. Child Development, 72, 506–517.
Newman, L. S., Cooper, J., & Ruble, D. N. (1995). Gender and computers: II. The
interactive effects of gender knowledge and constancy on gender-stereotyped
attitudes. Sex Roles, 33, 325–351.
Newman, M. A., Liss, M. B., & Sherman, F. (1983). Ethnic awareness in children: Not
a unitary concept. The Journal of Genetic Psychology, 143, 103–112.
Newman, L. S., & Ruble, D. N. (1988). Stability and change in self-understanding:
The early elementary school years. Early Child Development and Care, 40, 77–99.
Oaker, G., & Brown, R. (1986). Intergroup relations in a hospital setting: A further
test of social identity theory. Human Relations, 39, 767–778.
Ocampo, K. A., Bernal, M. E., & Knight, G. P. (1993). Gender, race, and ethnicity:
The sequencing of social constancies. In M. E. Bernal & G. P. Knight (Eds.),
Ethnic identity: Formation and transmission among Hispanics and other minorities
(pp. 11–59). New York: State University of New York Press.
Ogilvie, D. M., & Ashmore, R. D. (1991). Self-with-other representation as a unit
of analysis in self-concept research. In R. C. Curtis (Ed.), The relational self:
Theoretical convergences in psychoanalysis and social psychology (pp. 282–314). New
York: Guilford Press.
Ou, Y., & McAdoo, H. P. (1999). The ethnic socialization of Chinese American
children. In H. P. McAdoo (Ed.), Family ethnicity: Strength in diversity (2nd Ed.,
pp. 252–276). Thousand Oaks, CA: SAGE Publications.
Parke, R. D., & Buriel, R. (1998). Socialization in the family: Ethnic and ecological
perspectives. In W. Damon & N. Eisenberg (Eds.), Handbook of child psychology:
Vol. 3, Social emotional and personal development (pp. 553–617). New York:
John Wiley & Sons.
Perdue, C. W., Dovidio, J. F., Gurtman, M. B., & Tyler, R. B. (1990). Us and them:
Social categorization and the process of intergroup bias. Journal of Personality and
Social Psychology, 59, 475–486.
Phinney, J. S. (1989). Stages of ethnic identity development in minority group
adolescents. Journal of Early Adolescence, 9, 34–49.
Phinney, J. S. (1990). Ethnic identity in adolescents and adults: Review of research.
Psychological Bulletin, 108, 499–514.
Phinney, J. S. (1992). The multigroup ethnic identity measure: A new scale for use with
diverse groups. Journal of Adolescent Research, 7, 156–176.
Phinney, J. (1993). A three-stage model of ethnic identity development. In
M. Bernal & G. Knight (Eds.), Ethnic identity: Formation and transmission among
Hispanics and other minorities (pp. 61–79). Albany, NY: State University of New
York Press.
Phinney, J. (1996). When we talk about American ethnic groups: What do we mean?
American Psychologist, 51, 918–926.
Phinney, J. S., & Chavira, V. (1992). Ethnic identity and self-esteem: An exploratory
longitudinal study. Journal of Adolescence, 15, 271–281.
Powlishta, K. K. (1995). Gender bias in children’s perception of personality traits.
Sex Roles, 32, 17–28.
Quintana, S. M. (1998). Children’s developmental understanding of ethnicity and
race. Applied and Preventive Psychology, 7, 27–45.
Ramsey, P. G. (1991). The salience of race in young children growing up in an
all-white community. Journal of Educational Psychology, 83, 28–34.
 Ruble et al. 73
Ramsey, P. G., & Myers, L. C. (1990). Salience of race in young children’s cognitive,
affective, and behavioral responses to social environments. Journal of Applied
Developmental Psychology, 11, 49–67.
Rhee, E., & Ruble, D. N. (1997). Development of gender and racial constancy. Poster
session presented at the biannual meeting of the Society for Research in Child
Development, Washington, DC.
Rhee, E., Cameron, J. A., & Ruble, D. N. (2002). Development of gender and racial
constancy in European American and racial minority children. Manuscript in
preparation.
Rhee, E., Alvarez, J. M., McGlynn, E. A., & Mull, E. (2003a). Variation in European
American children’s racial attitudes and race-related perceptions and preferences.
Manuscript submitted for publication.
Rhee, E., Ruble D. N., Alvarez, J. M., Stangor, C., & Jones, J. M. (2003b). Relationship
between racial constancy and judgments about racial groups in African American,
Asian American, and European American children. Manuscript submitted for
publication.
Rholes, W. S., Newman, L. S., & Ruble, D. N. (1990). Understanding self and
other: Developmental and motivational aspects of perceiving people in terms
of invariant dispositions. In E. T. Higgins & R. Sorrentino (Eds.), Handbook
of motivation and cognition: Foundations of social behavior, Vol. II. New York:
Guilford Press.
Rowley, S. J., Sellers, R. M., Chavous, T. M., & Smith, M. A. (1998). The relationship
between racial identity and self-esteem in African American college and high school
students. Journal of Personality and Social Psychology, 74, 715–724.
Rubin, M., & Hewstone, M. (1998). Social identity theory’s self-esteem hypothesis:
A review and some suggestions for clarification. Personality and Social Psychology
Review, 2, 40–62.
Ruble, D. N. (1994). A phase model of transitions: Cognitive and motivational
consequences. In M. Zanna (Ed.), Advances in experimental social psychology,
Vol. 26, (pp. 163–214). New York: Academic Press.
Ruble, D. N., & Dweck, C. (1995). Self-conceptions, person conceptions, and their
development. In N. Eisenberg (Ed.), Review of personality and social psychology:
Social development, Vol. 15. Thousand Oaks, CA: SAGE Publications.
Ruble, D. N., & Frey, K. S. (1991). Changing patterns of comparative behavior as
skills are acquired: A functional model of self-evaluation. In J. Suls & T.A. Wills
(Eds.), Social comparison: Contemporary theory and research. Hillsdale, NJ:
Lawrence Erlbaum Associates, Inc.
Ruble, D. N., & Goodnow, J. (1998). Social development from a lifespan perspective.
In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), Handbook of social psychology. New
York: McGraw-Hill.
Ruble, D. N., Greulich, F., Pomerantz, E. M., & Gochberg, B. (1993). The role of
gender-related processes in the development of sex differences in self-evaluation
and depression. Journal of Affective Disorders, 29, 97–128.
Ruble, D. N., & Martin, C. L. (1998). Gender development. In D. W. Damon
(Ed.), Handbook of Child Psychology, Vol. 3 (5th ed.). New York: John Wiley &
Sons.
Ruble, D. N., & Ruble, T. L. (1982). Sex-role stereotypes. In A. G. Miller (Ed.), In the
eye of the beholder: Contemporary issues in stereotyping. New York: Holt, Rinehart,
& Winston.
74 Children’s collective identity
Ruble, D. N., & Seidman, E. (1996). Social transitions: Windows into social psycho-
logical processes. In E. T. Higgins & A. Kruglanski (Eds.), Handbook of social
processes. New York: Guilford Press.
Ruble, D. N., & Stangor, C. (1986). Stalking the elusive schema: Insights from
developmental and social analyses of gender schemas. Social Cognition, 4,
227–261.
Rutland, A. (1999). The development of national prejudice, in-group favouritism and
self-stereotypes in British children. British Journal of Social Psychology, 38, 55–70.
Sani, F., & Bennett, M. (2001). Contextual variability in young children’s gender
in-group stereotypes. Social Development, 10, 221–229.
Sellers, R. M., Rowley, S. A. J., Chavous, T. M., Shelton, J. N., & Smith, M. A. (1997).
Multidimensional inventory of black identity: A preliminary investigation of
reliability and construct validity. Journal of Personality and Social Psychology, 73,
805–815.
Sellers, R. M., Smith, M., Shelton, J. N., Rowley, S. A. J., & Chavous, T. M. (1998).
Multidimensional model of racial identity: A reconceptualization of African
American racial identity. Personality and Social Psychology Review, 2, 18–39.
Semaj, L. T. (1980). The development of racial evaluation and preference: A cognitive
approach. The Journal of Black Psychology, 6, 59–79.
Semaj, L. T. (1985). Afrikanity, cognition, and extended self-identity. In M. B.
Spencer, G. K. Brookins, & W. R. Allen (Eds.), Beginnings: The social and affective
development of black children (pp. 173–184). Hillsdale, NJ: Lawrence Erlbaum
Associates, Inc.
Serbin, L. A., Powlishta, K. K., & Gulko, J. (1993). The development of sex-typing in
middle childhood. Monographs of the Society for Research in Child Development,
58, Serial No. 232 (Whole issue).
Serbin, L. A., & Sprafkin, C. (1986). The salience of gender and the process of
sex-typing in three- to seven-year old children. Child Development, 57, 1188–1199.
Serbin, L. A., Sprafkin, C., Elman, M., & Doyle, A. B. (1984). The early development
of sex differentiated patterns and social influence. Canadian Journal of Social
Science, 14, 350–363.
Sigelman, C. K., & Singleton, L. C. (1986). Stigmatization in childhood: A survey of
developmental trends and issues. In G. Becker, L. M., Colemena, & S. Ainley (Eds.),
The dilemma of difference: A multidisciplinary view of stigma. New York: Plenum
Press.
Signorella, M. L., Bigler, R. S., & Liben, L. S. (1993). Developmental differences in
children’s gender schemata about others: A meta-analytic review. Developmental
Review, 13, 147–183.
Slaby, R. G., & Frey, K. S. (1975). Development of gender constancy and selective
attention to same-sex models. Child Development, 52, 849–856.
Smetana, J. G., & Letourneau, K. J. (1984). Development of gender constancy and
children’s sex-typed free play behavior. Developmental Psychology, 20, 691–695.
Spence, J. T., & Hall, S. K. (1996). Children’s gender-related self-perceptions, activity
preferences, and occupational stereotypes: A test of three models of gender con-
structs. Sex-Roles, 35, 659–692.
Spence, J. T., & Helmreich, R. (1978). Masculinity and femininity: Their psychological
dimensions, correlates, and antecedents. Austin, TX: University of Texas Press.
Spencer, M. B. (1982). Personal and group identity of black children: An alternative
synthesis. Genetic Psychology Monograph, 106, 59–84.
 Ruble et al. 75
Spencer, M. B. (1984). Black children’s race awareness, racial attitudes and self-
concept: A reinterpretation. Journal of Child Psychology and Psychiatry, 24,
433–441.
Spencer, M. B. (1985). Cultural cognition and social cognition as identity factors in
black children’s personal-social growth. In M. B. Spencer, G. K. Brookins, & W. R.
Allen (Eds.), Beginnings: The social and affective development of black children.
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Spencer, M. B. (1988). Self-concept development. In D. Slaughter (Ed.), Black
children and poverty: A developmental perspective (pp. 59–72). San Francisco:
Jossey-Bass.
Spencer, M. B., Brooklins, G. K., & Allen, W. R. (1985). Beginnings: The social and
affective development of black children. Hillsdale, NJ: Lawrence Erlbaum Associates,
Inc.
Spencer, M. B. & Markstrom-Adams, C. (1990). Identity processes among racial and
ethnic minority children in America. Child Development, 61, 290–310.
Stangor, C., & Ruble, D. N. (1987). Development of gender role knowledge and
gender constancy. In L. S. Liben & M. L. Signorella (Eds.), New directions for
child development: Vol. 39. Children’s gender schemata (pp. 5–22). San Francisco:
Jossey-Bass.
Steele, C. (1992). Race and the schooling of black Americans. The Atlantic Monthly,
269, 68–78.
Steele, C. M. (1997). A threat in the air: How stereotypes shape the intellectual iden-
tities and performance of women and African Americans. American Psychologist,
52, 613–629.
Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of
the self. Advances in Experimental Social Psychology, 21, 261–302.
Steele, C. M., Spencer, S., & Aronson, J. (in press). Stereotype threat. In M. Zanna
(Ed.), Advances in Experimental Social Psychology. New York: Academic
Press.
Stephan, W. G., & Rosenfield, D. (1979). Black self-rejection: Another look. Journal of
Educational Psychology, 71, 708–716.
Stipek, D., Gralinski, H., & Kopp, C. (1990). Self-concept development in the toddler
years. Developmental Psychology, 26, 972–977.
Szkrybalo, J., & Ruble, D. N. (1999). “God made me a girl”: Gender constancy
judgements and explanations revisited. Developmental Psychology, 35, 392–402.
Tajfel, H. (1978). Social categorization, social identity and social comparison. In H.
Tajfel (Ed.), Differentiation between social groups: Studies in the social psychology
of intergroup relations (pp. 61–76). London: Academic Press.
Tajfel, H. (1979). Individuals and groups in social psychology. British Journal of
Social Psychology, 18, 183–90.
Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In W.
Austin & S. Wochel (Eds.), The social psychology of intergroup relations (pp. 33–47).
Monterey, CA: Brooks/Cole.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In
S. Worchen & W. G. Austin (Eds), Psychology of intergroup relations (pp. 7–24).
Chicago, IL: Nelson Hall.
Taylor, S. E., Fiske, S. T., Etcoff, N. L., & Ruderman, A. J. (1978). Categorical and
contextual bases of person memory and stereotyping. Journal of Personality and
Social Psychology, 36, 778–793.
76 Children’s collective identity
Thoits, P. A., & Virshup, L. K. (1997). Me’s and we’s: Forms and functions of social
identities. In R. D. Ashmore & L. J. Jussim (Eds.), Self and identity: Fundamental
issues. Rutgers series on self and social identity, Vol. 1 (pp. 106–133). New York:
Oxford University Press.
Thompson, S. K. (1975). Gender labels and early sex-role development. Child
Development, 46, 339–347.
Turner, J. C. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel
(Ed.), Social identity and intergroup relations (pp. 15–40). Cambridge: Cambridge
University Press.
Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive
theory of group behaviour. In E. J. Lawler (Ed.), Advances in group processes, Vol. 2
(pp. 77–121). Greenwich, CT: JAI Press.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorization theory. Oxford: Basil Blackwell.
Van Ausdale, D. V., & Feagin, J. R. (1996) Using racial and ethnic concepts: The
critical case of very young children. American Sociological Review, 61, 779–793.
Vaughan, G. M. (1963). Concept formation and the development of ethnic awareness.
Journal of Genetic Psychology, 103, 93–103.
Wellman, H. (1990). The child’s theory of mind. Cambridge, MA: MIT Press.
Weinraub, M., Clemens, L. P., Sockloff, A., Ethridge, R., Gracely, E., & Myers, B.
(1984). The development of sex role stereotypes in the third year: Relationships to
gender labelling, gender identity, sex-typed toy preferences, and family character-
istics. Child Development, 55, 1493–1503.
Wigfield, A., & Eccles, J. S. (2000). Expectancy-value theory of achievement
motivation. Contemporary Educational Psychology, 25, 68–81.
Willer, S. C., & Romney, A. K. (1988). Systematic data collection. Newbury Park, CA:
Sage Publications.
Yee, M. D., & Brown, R. J. (1988). Children and social comparisons. Final report to the
ESRC, University of Kent.
Yee, M. D., & Brown, R. (1992). Self-evaluations and intergroup attitudes in children
aged three to nine. Child Development, 63, 619–629.
Yee, M., & Brown, R. (1994). The development of gender differentiation in young
children. British Journal of Social Psychology, 33, 183–196.
Yip, T., & Fuligni, A. J. (2002). Daily variation in ethnic identity, ethnic behaviors, and
psychological well-being among American adolescents of Chinese descent. Child
Development, 73, 1557–1572.
3 Developmental aspects of
social identity
Fabio Sani and Mark Bennett

Our interest is in the development of social identities and children’s under-

standing of those identities. In this chapter we present a series of studies
emerging from this interest, studies that are part of a wider research pro-
gramme on developmental aspects of self and identity. To start with, we
discuss qualitative developmental change in children’s conception of norma-
tive features of members of social groups. Essentially, the question we try to
address is: Can we assume that, when thinking about the normative features
of group members, children focus on the same aspects of their identities as
adults do? Social identity and self-categorisation theorists tend to equate, at
least implicitly, normative aspects of group members with the beliefs, values,
and attitudes of group members (e.g., Bar-Tal, 1998). However, we will sug-
gest that children’s understanding of normative aspects of group members,
and therefore children’s knowledge of group identities, moves through age-
related stages, and that a focus on identity-related beliefs and attitudes may
be a relatively late achievement.
Following this, we investigate whether children’s representation of the con-
tent of in-group identity varies according to changes within the intergroup
context in which an identity is construed, and when this starts to happen. In
particular, we consider the possible contextual variability of both national
and gender identities. The likelihood that there may be variations in con-
ceptualisations of identities as a function of the comparative context has
important implications for the way in which children’s stereotypes are under-
stood; indeed, it may pose a challenge to widely held cognitive-structural
views of stereotypes.
Finally, we discuss developmental aspects of the inclusion of a group in the
self. Thus, the question we address is: When do children start subjectively
identifying with social groups? When do they intentionally embrace the
group norms and represent the self in a stereotypical fashion (i.e., in terms of
the group prototype)? In sum, when do children start behaving, thinking, and
feeling in terms of a collective self ?
We conclude by suggesting that there are many specifically developmental
issues that need to be addressed by the social identity approach. In particular,
we make the case that more attention should be directed to the particular
78 Developmental aspects of social identity
difficulties associated with the investigation of children’s subjective identifica-
tion with the group. Finally, we note that although social identity theory and
self-categorisation theory have considerable heuristic potential for develop-
mentalists, it is nonetheless important to observe that, as theories proposed
for adult behaviour, they make a variety of assumptions that may constrain
their developmental application.

CHILDREN’S CONCEPTION OF SOCIAL IDENTITIES

People’s understanding of the normative aspects of group members’ iden-

tities can be based on many types of information. For instance, ethnicity can
be conceived in terms of skin colour, facial features, appearance based on
dress, psychological properties, values, beliefs, and culture. Clearly, adults
may draw upon any or indeed all of these different characteristics in con-
ceptualising social groups. However, we suggest that children’s conceptualisa-
tions of the normative features of group members are likely initially to be
quite limited and may undergo substantial elaboration with increasing age.
Specifically, we suggest that young children may conceive of group members’
normative features in terms of actions and dispositions; only later might
they also consider identity-related beliefs. Our hypotheses are based upon
work concerning children’s person perception (including self-perception) and
explanations for interpersonal events.
First of all, in their descriptions of self and other, although there is much
evidence that young children are interested in actions (Bartsch & Wellman,
1995) and dispositions (e.g., Damon & Hart, 1988; Eder, 1990), the same is
not true with respect to identity-relevant beliefs (e.g., “I believe in world
peace because I don’t think wars solve anything”; from Damon & Hart, 1988,
p. 69). References to such beliefs typically do not appear until late childhood
or adolescence. Considered together, these findings are perhaps unsurprising
in the sense that where actions and dispositions may have a reasonable
degree of predictive utility in terms of future behaviour, the same is not true
with respect to identity-related beliefs. As such, it is understandable that,
developmentally, priority is given to actions and dispositions.
Going further, we propose that during middle to late childhood there may
be a progression from an individualistic to a more collective conception of
the normative features of group members that also recognises the role of
supra-individual phenomena, such as shared values and belief systems. In a
pilot study of children’s conceptions of their national identity (Sani, Lyons,
& Barrett, 1996), we found evidence broadly consistent with this proposal,
such that there was a move from conceptions that emphasised dispositional
and concrete aspects of national identity (“English people like football”;
“Scotland has nice countryside”) to those that emphasised belief-based and
collective aspects of national identity (“We believe in democracy”; “We value
fair play”). Lambert and Klineberg (1967), too, looking at conceptions of
 Sani and Bennett 79
nationality, found that it was not until late childhood that references were
made to political and religious belief. Such findings, coupled with the general
observation of a move in children’s cognition from the concrete to the
abstract, suggest the proposal that in representing normative features of
group members, young children are likely to focus on features such as
behavioural, physical, and dispositional attributes (which have concrete and
specifiable manifestations). Only later might they start conceiving of
normative features in terms of socially shared beliefs and values (which are
abstract and lacking clear behavioural implications). In order to examine this
proposal, we conducted three separate studies, which we now outline.

Study 1: Children’s understanding of normative features in the

context of the Northern Irish conflict
Our first study (Sani, Bennett, Agostini, Malucchi, & Ferguson, 2000; Study
1) looked at Northern Irish children’s explanations for the conflict that has
long existed between Catholic and Protestant communities in Northern
Ireland. What we were interested in was whether, in accounting for group
members’ behaviour, children would appeal to intrapsychic dispositions
(“They’re mean and like to start a fight”) or to socially shared values and
beliefs (“Catholics won’t give up their belief that it should just be one coun-
try”). We decided to use an intergroup situation as the stimulus for the pro-
duction of behavioural explanations because, in accordance with the social
identity approach, we assume that human groups and their relevant features
become cognitively salient and meaningful through intergroup comparisons
(Tajfel, 1981; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). The North-
ern Irish context was selected because the serious and chronic form of con-
flict found in that province is likely to provide a high level of experience of
issues of group membership; as such, it provides a strict test of our proposals
concerning cognitive-developmental limitations, in that “poor” performance
might not be readily ascribed to a lack of relevant experience.
Six-, nine-, and twelve-year-olds (with approximately equal numbers of
boys and girls and Roman Catholics and Protestants at each age level) were
asked semistructured questions that focused upon the relationship between
Catholics and Protestants in Northern Ireland. Specifically, the children were
first asked: “Do you think there are any problems between Catholics and Prot-
estants in this country?” In case of affirmative answers, they were then asked,
“What do you think causes the problems?”
The 12-year-old children were found to be significantly more inclined than
the 6- and 9-year-olds to acknowledge the conflict between the two religious
groups: Nearly all the oldest children recognised the conflict, but surprisingly
only about half of the two younger groups did so.
Concerning the causes of conflict, we found a very strong association
between age group and type of explanation provided, which provided evi-
dence of a progression from a tendency to use individualistic explanations to
80 Developmental aspects of social identity
a willingness to describe the genesis of the events in terms of collective beliefs.
The younger children had a propensity to use individualistic explanations, the
9-year-olds could be viewed as transitional, in that some already used belief-
based accounts, (though many still employed individualistic accounts),
whereas nearly all 12-year-olds explained category members’ actions in terms
of supra-individual features.
The unexpected feature of the data was that so many children, particularly
the younger ones, failed to acknowledge the existence of conflict between the
communities of Northern Ireland. This could reflect a general failure to
understand conflict at the level of relatively abstract social groups such as
those studied here; alternatively it may be the result of parents’ and teachers’
efforts to shield children from the conflict. Regardless of why these findings
emerged, the data relevant to our hypothesis about developments in child-
ren’s thinking in this domain were limited by the fact that sample sizes were
substantially reduced; moreover, in looking only at children who were aware
of the conflict, we may have had an atypical subsample. In order to address
these problems, we conducted a further study in which we ensured that all
participants possessed requisite knowledge about the existence of a specified
conflict.

Study 2: Children’s understanding of normative features in the

context of a fictional intergroup conflict.
This study (Sani et al., 2000; study 2) involved 8-, 10-, and 12-year-old Italian
children. Unfortunately, 5-year-old children were not recruited because in the
course of a pilot study it became clear that many of them were unable to cope
with the memory demands of the task.
Children were shown a story in the form of cartoons. At the beginning of
the story a specific group, called “People of the Mountains”, was presented.
Then, information about (1) external, physical characteristics, (2) psycho-
logical/behavioural characteristics, and (3) socially shared beliefs of the
group members was provided. For each of the foregoing three aspects, two
pieces of information – one conflict-relevant and one conflict-irrelevant –
were provided. (The purpose of the inclusion of both relevant and irrelevant
information for each of the three aspects was to ensure that children’s prefer-
ence for specific levels of explanations over others reflected a proper under-
standing of the relation between particular antecedents and consequences,
rather than a general preference for one level of explanation over others.) So,
for instance, concerning psychological/behavioural characteristics, it was
stated that the members of one group were “rather aggressive” (relevant
information) and that “they tend to be very quiet” (irrelevant information).
Finally, a social conflict with the out-group (the “People of the Valley”) was
described: it was specified that the People of the Mountains invaded the
village in the plains surrounding the mountains, where the People of the
Valley lived, and forced them to worship their gods in the temple.
 Sani and Bennett 81
We found significant associations between age and types of explanation
provided for the conflict. Explanations based on beliefs systems were used by
one third of the 12-year-old children, but by very few 10- and 8-year-olds. On
the other hand, the majority of 10- and 8-year-olds used explanations based
on psychological/behavioural characteristics; these groups invoked such
explanations significantly more frequently than did 12-year-olds. (None of
the children used explanations that included conflict-irrelevant information.
Therefore, within a particular level of explanation, it appears that children’s
selection of particular explanations is appropriate.)
As in the previous study, there was evidence of a progression from explan-
ations based on the individualistic features of group members to explanations
based on supra-individual aspects. However, it is noteworthy that the overall
proportion of the oldest children referring to belief systems was markedly
lower in this study than in the initial study (34% vs. 94%). This discrepancy
between the studies may reflect the fact that the initial study, unlike the present
study, involved genuine conflict about which, over a period of years, the older
children were likely to have gathered a wealth of information from many
diverse sources (Cairns, 1990). More generally, we should also note that, over
both studies, explanation in terms of shared belief emerged slightly later than
we had predicted. Conceivably this may have been due to a preference for
explaining aggressive behaviour in terms of corresponding dispositions,
rather than to a lack of understanding of belief-based aspects of behaviour.
Thus, in the study that follows, we move away from children’s explanations
for conflict and instead rely on direct questioning about social identities.

Study 3: Children’s descriptions of group identities

In the studies described thus far we investigated how children account for
group members’ behaviour and looked in particular at their explanations for
intergroup conflict. Although the data were broadly consistent with our pre-
dictions, a weakness of these studies, which may have given rise to “noisy”
data, is that they tackled the issue of children’s conception of the normative
features of social group members in an oblique manner: Rather than asking
directly about the characteristics of group members, such information was
sought in the explanations offered for conflict. Thus, Study 3, using direct
questioning, aimed to cast light on the way in which children and adults
represent the identity of several types of group, including both in-groups and
out-groups.
This study involved more than 250 Scottish participants: 5-, 8-, 11-, and 14-
year-olds, and adults. Initially participants were presented with cards, each
with a descriptor of a particular social identity (e.g., child, adult, Christian,
Muslim, Scottish, English, Girl Guide, Cub, Socialist, Scottish Nationalist,
boy, girl, white, black, etc.), and were asked to select those identities that
described themselves. The selected cards were then rank-ordered in terms of
their subjective importance. Finally, participants were required to consider
82 Developmental aspects of social identity
their two most important social identities and in randomised order, and for
each identity, were asked: “What are [boys/children/Scottish people/etc.]
like?”; “What do [. . .] do?”; “What do [. . .] want?”; “What do [. . .] believe?”.
Thus, we asked a broad range of open-ended questions in order to generate a
data set that might reflect participants’ conceptions of social identities. In
addition, and in order to address the possibility that children may have know-
ledge of belief-based groups but not judge those groups as subjectively
important (and thus not be asked about them), we also asked about two such
groups likely to be familiar to them: Christians and Scottish Nationalists.
(The research was conducted at the same time as the election of Members
of the Scottish Parliament (MSPs) to the then newly established Scottish
Parliament and debates concerning Scottish Nationalism were conspicuous
and vigorous at the time.)
Responses were content-analysed. The unit of analysis was the response to
each specific question asked. So, for instance, the answer to the question
“What are [. . ..] like?” constituted a unit to be allocated to a category. Six
theoretically relevant categories were used in coding responses to each of the
questions. The first category was concerned with behavioural/psychological
attributes (e.g., “kind”, “likes sharing”, “don’t like boys”, “go shopping”)
that were judged by coders to be identity-relevant for the members of that
particular group. For example, if boys were described as liking football, fight-
ing, and fishing, this was considered as identity-relevant for boys. The second
category was concerned with behavioural/psychological attributes that are
irrelevant to identity. So, for instance, responses in which boys were described
as friendly or pet-loving were allocated to this category; that is, while these
attributes may be true of particular cases, they are not consistent with gen-
eral, category-based expectations in this instance. The third category included
responses based on beliefs that are relevant but not defining for the group that
is described. For example, when Christians were described as “charitable”
this may be seen as a relevant belief for this group, but not as defining. The
fourth category was concerned with responses based on beliefs that are both
relevant and defining for a given group. For example, when Christians were
described as believing in God, it is clear that this description is concerned
with something that is central for members of the Christian group (but it
would not be crucial for members of the groups of children, Socialists, and
whites). The fifth category concerned responses based on beliefs that are
irrelevant to the group to be judged (and so cannot be defining either).
Finally, category six was a residual category, other.
The data set is extremely rich and many analyses remain to be conducted.
However, our hypotheses were largely supported. Young children’s concep-
tions of social identities were principally concerned with personal and
behavioural attributes; little reference was made to belief-based attributes.
Among the youngest children, preferred identities related to gender and age
and were seen as characterised by behavioural and dispositional features (e.g.,
“boys like to play football”). Only among the older children, adolescents,
 Sani and Bennett 83
and adults was there a clear recognition of the role of beliefs in some social
identities (e.g., “to be a Christian means believing in God”, “the nationalists
want independence for Scotland”). This is not to say that young children
entirely failed to understand belief-based identities; rather, their view is par-
tial. For example, younger children recognised that Scottish Nationalists hold
very positive views of Scotland, but failed to provide evidence of understanding
that they are defined by their beliefs concerning Scottish independence.
Looking at the data in more detail, what emerges is that all age groups
made reference to identity-relevant psychological attributes, so that even
5-year-olds identified at least one relevant attribute per group considered
(for example, commenting that, “girls like going shopping”). Eight-, eleven-,
and fourteen-year-olds identified significantly more such attributes (typically
around 1.6 per group). Adults on average noted more still (around 2 identity-
relevant psychological attributes per group).
Turning to identity-defining beliefs, 5-year-olds, as we had predicted,
made no reference to such attributes. Eight-, eleven-, and fourteen-year-olds
made significantly more references to them and did not differ from each other
with, on average, a total of 1.2 such attributes over the four groups con-
sidered. Adults offered significantly more examples than did 8-year-olds,
typically 1.5 identity-defining beliefs over the four groups considered.
Although the age differences are broadly as we had predicted, it is noteworthy
that when directly asked about beliefs, even 8-year-olds provide evidence of
comprehension, suggesting that our previous method underestimated ability
in this respect. It is also noteworthy that even among adults the mean number
of references to social identity-defining beliefs was quite small. The fact that
means for behavioural/psychological attributes were larger than those for
beliefs probably reflects the fact that while all social identities are associated
with the former, not all are associated with the latter. For example, the groups
“teenagers”, “children”, “boys”, “girls”, “whites”, “blacks”, and “Scottish”
(to name but a few) have no identity-defining beliefs associated with them.
Thus, differences over the coding categories considered here need to be
viewed with this caution in mind.
A nearly identical pattern of findings emerged for identity-relevant beliefs
as for identity-defining beliefs. Thus, means for all groups were relatively low
compared to those for behavioural/psychological attributes, but predicted age
trends nonetheless emerged.
Finally, looking at identity-irrelevant references to both behavioural/
psychological attributes and beliefs, what is striking is the absence of age
differences. We had expected that younger children’s conceptions of social
identities would be “noisy”, incorporating information that both agreed with
and diverged from the adult conception. As we saw it, children’s develop-
mental task would be both to gather relevant information and to sift out
irrelevant information. Our data, however, appear not to support this pro-
posal. In this respect, the difference between young children and adults would
appear to be in the extent, rather than the clarity, of their knowledge about
84 Developmental aspects of social identity
social identities, suggesting that from a very early age children may be
especially sensitive to identity-relevant information – indeed, a claim central
to schematic models of identity, such as gender-schematic processing theory
(Martin & Halverson 1981).

Summary
Overall, what emerges from these studies is that, as predicted, younger chil-
dren conceive of social identities primarily in terms of group members’
behavioural and dispositional attributes. Not until later do they come to
recognise the importance (for some identities at least) of shared beliefs and
values1. This is not to suggest that there are no important developments prior
to the age of 5 years. Indeed, Quintana (1998) has suggested that among
3- to 5-year-old children, ethnic groups may be conceived primarily in
physical terms (e.g., with reference to skin tone, facial morphology, etc.).
Such a tendency may be found for other types of group, too. Regardless of
such a possibility, our findings suggest that belief-based forms of social iden-
tity may not be fully grasped until mid to late childhood.
For many identities, it is plainly inconsequential that belief-based attributes
are not understood. For example, gender identities appear to be based primar-
ily on behavioural and dispositional attributes; there are no beliefs that can be
taken as necessary and sufficient for gender group membership. The same is
self-evidently not the case with respect to identities such as “Christian”,
“Muslim”, “Socialist”, and “animal rights activist”. For such identities,
beliefs play a definitional role. Our assumption in this chapter has been that
owing to cognitive-developmental limitations, children experience difficulties
in conceiving of abstract entities such as socially shared beliefs. However, we
accept that a further explanation for this progression may be that belief-based
identities involve personal choice; that is, they reflect agents’ commitments.
Interestingly, the sorts of identities children understand well (e.g., gender
identity, age-based identity, ethnic identity, etc.) are de facto identities –
identities about which they have little choice. To the extent that such identities
are largely immutable and are significant across diverse social contexts,
we speculate that parents, in socialising children, may give these identities
priority; ascribed identities, virtually by definition, may be “scaffolded” to a
greater degree than those reflecting personal choice. With this in mind,
we recognise the possibility that the developmental progression we have

1 Theory of mind researchers may find this puzzling. After all, a wealth of evidence now shows
that even 4-year-olds understand belief (e.g., see Perner, 1999). However, it is important to
point out that the sorts of beliefs considered here differ from those investigated by theory of
mind researchers. For example, in the latter tradition beliefs are empirical in nature and involve
a world-to-mind “direction of fit” (Searle, 1983); the beliefs considered here are essentially
ideological and cannot be seen as driven by the empirical facts of the world.
 Sani and Bennett 85
identified may not primarily be a cognitive-developmental phenomenon, but
instead one based on domain-related increments in expertise (Chi & Rees,
1983). It remains for future research to address the relative contributions of
experience and cognitive development in children’s understanding of social
identities.
Finally, then, a general point to note is that social identities differ. This is
not something that has been widely addressed by the social identity
approach. Deaux, Reid, Mizrahi, and Cotting (1999) have speculated that
“this neglect reflects a concern with general psychological processes that are
assumed to characterize a variety of specific forms of identification” (p. 93).
From a developmental perspective, we suggest that the evidence provided by
our work points to a need to examine diverse types of social identity,
the challenges of conceptualisation that they pose for children, and the
challenges of socialisation that they pose for adults.

THE FLEXIBILITY OF CHILDREN’S UNDERSTANDING OF

IN-GROUP IDENTITY

Thus far we have considered general developmental change in the conceptual-

isation of social identity. However, it is important, too, to consider possible
contextual variation in this domain. That is, to what extent are children’s
conceptions of particular groups context-specific? In what follows we shall
introduce some of our own studies, which suggest that children, like adults,
may judge a given group on the basis of the particular intergroup situation in
which a judgment is made. First, however, to place this work within a broader
theoretical context, we turn to social psychological work demonstrating that
adults’ conceptions of social identities (as expressed in stereotypes) are
subject to contextual variation.
Within contemporary social psychology, a stereotype is widely conceived
as a set of beliefs about attributes and behaviours of members of a social
category (Hamilton & Trolier, 1986; Stangor & Shaller, 1996). Stereotypes are
viewed as stored in memory, and automatically assigned to people who are
categorised as members of the social category associated to those specific
stereotypes (e.g., Fiske & Neuberg, 1990). Clearly, this view implies a con-
ceptualisation of stereotypes as essentially fixed and rigid mental structures,
which, as asserted by Stangor (1995, p. 631), may change only through inter-
action with individuals from the stereotyped group, which disconfirm existing
stereotypes, or because the perceiver does not consider the stereotype as
useful or desirable any longer.
An obvious implication of this position is that the content of a stereotype
is unaffected by the nature of the intergroup context within which the stereo-
type is invoked. So, for instance, according to this position my conception of
psychologists in terms of, say, their methodological orientation (e.g., whether
they are rigorous, objective, imaginative, etc.) will remain the same, regardless
86 Developmental aspects of social identity
of whether I am thinking about psychologists in comparison to historians or
whether I am thinking about them in comparison to chemists.
More recently, this cognitive-structural view of stereotypes has been
criticised by self-categorisation theorists (Oakes, Haslam, & Turner, 1994),
who have rejected the notion that a stereotype is a stored concept “waiting-
to-be-activated”. They contend that (1) a stereotype is always formed within a
specific intergroup context, either explicit or implicit, as stereotypes are used
to explain, describe, and justify the nature of intergroup relations, and that
(2) in forming stereotypes, people make use of all cognitive resources that are
available to them. This implies that a change in the intergroup context within
which a given stereotype is used is likely to lead to a change in the nature of
the stereotype itself. Therefore, the substantive content of a stereotype
(regardless of whether the stereotyped group is an in-group or an out-group)
is inherently comparative, flexible, and variable.
However, self-categorisation theorists do not just emphasise stereotype vari-
ability; they also contend that the different understandings of a certain group
that may arise in different intergroup contexts arise systematically, and not
merely haphazardly. It is argued that in each particular context a group’s defin-
ition is based on an accentuation of its differences from the other groups on
relevant dimensions of comparison. The function of this process of accentu-
ation is that of allowing the perceiver to structure the social environment and to
make sense of the situation (Oakes et al., 1994; Turner & Oakes, 1997).
Self-categorisation theorists have supported their position with strong evi-
dence. For instance, Haslam, Turner, Oakes, McGarty, and Hayes (1992) found
that the way in which Australians described Americans during the Gulf War
changed as a function of the number and type of nations included in the com-
parative context, and that changes occurred particularly on dimensions that
were relevant to the ongoing military conflict. Thus, Americans were charac-
terised as more “aggressive” when the context included the USSR than when
it did not, indicating that subjects perceived differences between America and
the USSR in their approach to the war and to East–West relations in general.
These general findings, which have been replicated by other studies looking
at further identities (e.g., Haslam & Turner, 1992; Hopkins & Murdoch, 1999;
Hopkins, Regan, & Abell, 1997; Sani & Thomson, 2001), have important
implications for developmental studies on stereotypes, as developmental psy-
chologists seem to subscribe, at least implicitly, to a view of stereotypes as
endemically rigid and fixed. Researchers have assumed that at any point in
development children’s understanding of social categories is based on a spe-
cific and cross-situationally stable stereotypical content, which may change
only through more general social-cognitive development. Thus, attempts have
been made to specify, at different points in development, children’s stereo-
types of national groups (Barrett & Short, 1992; Lambert & Klineberg,
1967), gender groups (Best et al., 1977; Serbin & Sprafkin, 1986), age groups
(Edwards, 1984), and ethnic groups (Davey, 1983). However, results emerging
from the research conducted by self-categorisation theorists suggest the
 Sani and Bennett 87
desirability of investigating the issue of stereotype variability from a
developmental perspective.
From the perspective of the social identity approach, the process of stereo-
typing clearly rests on a capacity to engage in social comparison. That is,
“we” may seem easygoing when compared to Group X, but when compared
to Group Y, may appear quite formal. Interestingly, Ruble and her colleagues
(Pomerantz, Ruble, Frey, & Greulich, 1995; Ruble, 1983; Ruble & Flett, 1988;
Ruble & Frey, 1991) have found that only at around 7 years old do children
draw upon social comparison information in order to make judgments about
the self. Thus, contextual variability of stereotype content may be a relatively
late development.
A further developmental consideration is that young children have relatively
compartmentalised notions about the self, reflecting “a rudimentary ability to
intercoordinate concepts” about the self (Harter, 1999, p. 41). Only by mid
childhood does there appear to be the sort of cognitive flexibility in self-
concepts that is presupposed by the account of stereotyping presented by
self-categorisation theory. In the light of such considerations, we assume that
contextual variability in stereotyping is likely to emerge in children aged around
7 years. Clearly, this assumption has important implications for the develop-
mental study of social identity. As we have already mentioned, in-group identity
can be legitimately seen as based on a stereotypical representation of in-group
members, including self. As a result, once applied to the in-group, stereotype
variability is conceptually equivalent to social identity variability.
In an initial study, we investigated children’s national stereotypes (Sani,
Bennett, & Joyner, 1999), adapting the procedure used by social psychologists
(e.g., Haslam et al., 1992). That is, we solicited stereotypes of the national in-
group following consideration of two distinct out-groups (Spanish/
Germans). Contrary to expectation, in-group stereotypes were unaffected by
the comparative context. That is, children described the in-group in the same
way irrespective of whether it was described following consideration of Span-
ish people or of German people. (Interestingly, Barrett, Wilson, & Lyons,
(2003), using a different methodology, obtained similar findings to ours.)
Conceivably, however, rather than reflecting essentially fixed stereotypes,
these results may have been due to two main weaknesses of the study. First,
this study was concerned with an aspect of social identity, that is, nationality,
which may not be salient for many young children (Bennett, Lyons, Sani, &
Barrett, 1998a). Second, since children were presented with a brief and ready-
made (rather than solicited) description of the out-group, this procedure may
have failed to establish a comparative context. These considerations formed a
basis for the design of Study 4.

Study 4: On the flexibility of children’s gender identity

In this study (Sani & Bennett, 2001), we explored possible contextual vari-
ability in 6- and 7-year-old children’s gender stereotypes, using a simple trait
88 Developmental aspects of social identity
selection procedure (i.e., children were required to select from a large set of
cards those which best described a specified group) as in the Sani et al. (1999)
study. Two experimental conditions were employed. In the first condition the
out-group was represented by adult “men” if participants were boys, or by
adult “women” if participants were girls. In the second condition the out-
group was represented by “girls” if participants were boys, or by “boys” if
participants were girls. Obviously, in both conditions, the in-group was either
“girls” or “boys”, depending on the participants’ sex.
Importantly, there were clear indications that some stereotypical traits
attributed to the in-group changed significantly as the frame of reference
changed. Girls selected “friendly” more often when describing the in-group
after describing boys than after describing adult women, and boys selected
“strong” and “brave” more often when stereotyping the in-group after
describing girls than after describing adult men. Moreover, consistent with
SCT, the dimensions that varied according to changes in the comparative
context were those dimensions that seemed to maximise perceived intergroup
difference.
However, despite being consistent with predictions, the effects of the
comparative context were not as strong as expected: Only 2 adjectives used
by males and 1 used by girls, out of 27 adjectives that each group con-
sidered, varied cross-situationally. This may have reflected methodological
limitations of the study. In particular, we suspect that adjective selection
was an unduly “noisy” measure. First, when children are presented with a
large number of adjectives, they might make their selections based not on a
consideration of the entire set, but upon those adjectives that are noticed
either early or late in the task. Second, children may tend to select positive
and highly accessible adjectives, for instance “friendly”, “happy”, and
“nice”, at the expense of more neutral and slightly less accessible adjectives
such as “greedy”, “talkative”, and “loud”, whatever the nature of the inter-
group context. That implies that, if some of the more neutral or slightly
negative adjectives were seen as characterising many in-group members in a
given intergroup context, but only some in-group members in a different
intergroup context, predicted effects would be hard to observe. Finally, we
should note that this study looked only at an age group that had been
predicted to show the predicted effects of comparative context on stereo-
type content. In the following study, younger children are included in the
sample in order to address the hypothesis that their responses would not
show the predicted effects of comparative context. Moreover, a different
method was employed.

Study 5: A further investigation into the flexibility of children’s

gender identity
Study 5 (Sani, Bennett, Mullally, & MacPherson, 2003), using a sample of
5- and 7-year-old children, retained the same design as Study 4 but addressed
 Sani and Bennett 89
the limitations of that study. First of all, we reduced the number of adjectives,
selecting 18 of the original 27. (We discarded those adjectives that were never
used by more than one child in any of the experimental conditions in Study
4.) Second, we investigated children’s stereotypes by means of a rating task.
Children were asked to look at each card in the set and decide whether the
adjective written on the card was applicable to most girls/boys/grown-up
women/grown-up men, some, or none.
Results provided clear evidence that 7-year-old children conceive the in-
group identity in a flexible fashion. Boys’ conceptions of the in-group were
more likely to draw attention to being “brave”, “big”, and “strong” in the
context of girls than of men, but “talkative” in the context of men, rather
than girls. Girls’ in-group descriptions emphasised cleverness and hardwork-
ingness more strongly in the context of boys than of women, but greediness,
loudness, and smallness in the context of women, rather than boys. Therefore,
results seem to show that a stereotype is not an unchanging representation of
a social group but a representation of a group in relation to one or more
other groups. Also, results suggest that the adjectives that are rated in differ-
ent ways according to changes in the comparative context are those that
maximise perceived intergroup differences. So, this is why boys indicated, for
instance, that the adjectives “clean”, “happy”, and “kind” are applicable to
the same number of in-group members in both conditions, but that “brave”,
“big”, and “strong” are applicable to a higher number of in-group members
in comparison to girls than in comparison to men. Clearly, reference to such
qualities in the context of girls is a way to accentuate aspects of the in-group
that can be used to differentiate it from the relevant out-group on specific
dimensions.
Contrary to our initial predictions, even among the 5-year-olds there
appears to be some degree of flexibility in the construction of their gender
identity, though only among boys. Thus, boys described in-group members
as being more dirty and loud in the context of men than in the context of
girls, and as being tougher in the context of girls than in the context of men.
Again, the dimensions that varied according to changes in the intergroup
context appear to be those dimensions that maximise perceived differences
between the two groups. This indicates that 5-year-olds may possess, at least
in rudimentary form, the cognitive competencies necessary to demonstrate
sensitivity to comparative context2.

2 A puzzling aspect of the data is the gender difference observed among 5-year-olds in the
present study. An explanation for this difference may be that although gender is a highly
significant aspect of both boys’ and girls’ identities (Eckes & Trautner, 2000), socialisation
agents’ gender-related expectations of boys are more sharply defined than those of girls (Fagot
& Hagan, 1991). As a consequence, boys may have a more coherent set of gender-related self-
beliefs than girls, and may be more motivated to affirm these self-beliefs when their gender
identity is made salient. Consistent with this is Premack and Premack’s (1995) contention that
boys may be more sensitive than girls to group-based distinctions.
90 Developmental aspects of social identity
Summary
The results of our studies on gender identity (though not national identity)
demonstrate that conceptions of in-group identity vary as a function of the
comparative context. In characterising gender in-group identity, children do
not appeal to identical descriptors regardless of intergroup context. These
results are consistent with Banerjee and Lintern’s (2000) finding that 4- to 6-
year-old boys’ self-descriptions were significantly more gender-stereotypical
when before a group of same-sex peers than when alone.
These findings are surprising insofar as they indicate that even young
children provide evidence of sensitivity to comparative context, contrary to
what we had predicted. Thus, 5-year-olds appear to possess, to some degree,
the cognitive competencies necessary to demonstrate sensitivity to compara-
tive context, at least in the case of gender identity. That such competencies
are not seen in the domain of national identity is perhaps unsurprising. First
of all, as a highly abstract and inclusive category, nationality has significantly
lower utility than does gender. Whereas gender is an effective basis for
distinguishing between people within one’s typical social contexts, the same
is not true of nationality since the overwhelming majority of one’s peers,
relatives, teachers, and so on are likely to be of the same nationality as one-
self. Given its relatively low utility, nationality will be less salient than gender,
and in consequence, less is likely to be known about nationality-appropriate
characteristics than about gender-appropriate characteristics. Thus, although
young children appear to know the national labels (e.g., American, Italian;
see Barrett, Lyons, & del Valle, Chapter 6), they do not yet have the wealth of
knowledge about the content of the categories to provide a basis for the
metacontrasts assumed by self-categorisation theory.
A general question raised by our results is whether children’s representa-
tions of stereotypes, and group identities in general, can be considered as
stored concepts waiting to be activated, as is widespread within develop-
mental psychology. Researchers in the social identity tradition propose two
different responses to the issue. On the one hand, Turner and Onorato (1999)
suggest that in-group stereotypes are “created as they are used, on the spot,
and brought into being as they are brought into sight” (p. 32). Although
these authors do not deny that knowledge that is used to construe specific
stereotypes is stored in our cognitive system, they do not believe that this
knowledge is stored in the same form as the stereotype that is phenomenally
experienced. Instead, it exists as complex theoretical knowledge about
ourselves and the world in which we live, and is used flexibly to create specific
in-group stereotypes on the basis of current needs, expectations, and
sociostructural constraints.
In opposition to this view, Hogg (2001) contends that people have a stored
set of group stereotypes (including stereotypes of in-groups) that are situ-
ationally adjusted. That is, stereotypes are not created on the spot, but are
carried in people’s heads and modified at the moment in which they are
 Sani and Bennett 91
employed on the basis of the nature of the intergroup context. Clearly, this is
a key issue inasmuch as it speaks to the question of whether we should see
stereotypes and identities essentially as structures or as processes. (See David,
Grace, & Ryan, Chapter 5, for a fuller discussion of this matter.) Further-
more, it is an issue that is amenable to empirical inquiry. But as a subtle and
complex matter, and one that will mobilise strong (and competing) theor-
etical commitments, it is unlikely to see an easy resolution. Regardless of such
grand-scale debates, an important focus of inquiry should be the further
investigation of the social and developmental factors affecting the extent of
contextual variability in children’s identities.

THE EMERGENCE OF SOCIAL IDENTITY

Irrespective of developmental and contextual variations in children’s con-

ceptions of social identity, a key issue that remains is the matter of the
internalisation of social identities. That is, children may have knowledge of
identities, and may use that knowledge flexibly, but they may not subjectively
experience themselves as members of particular collectivities. Thus, as Ruble
et al. (Chapter 2) put it, we need to address “the developmental course of
children’s sense of ‘we’ ”.
Previous developmental researchers have often equated the social self with
the description of oneself in terms of particular social categories (“I am a
girl”; “I am a boy scout”; “I’m Scottish”, etc.). Although this is an important
first step in the development of social identity, it is no more than that: a first
step. We would make the case that merely describing oneself in terms of
social categories is not equivalent to social identity as conceived within the
social identity approach. As Turner, Oakes Haslam, and McGarty (1994)
point out, “social identity can be distinguished from personal identity in
terms of self-referential cognitions that identify ‘we’ and ‘us’ rather than ‘I’
and ‘me’ ” (p. 454). Among adults, it is reasonable to assume that social
self-descriptions, such as Democrat, Catholic, or whatever, imply subjective
identification with the particular collectivity. Indeed, much research provides
evidence to this effect (e.g., see Abrams & Hogg, 2001; Tyler, Kramer, & John,
1999). However, we suggest that in a developmental context it becomes
important to question whether children’s early self-referential descriptions
such as “I’m a boy” and “I’m Italian” connote that sense of “we” and “us”
that properly constitute social identity. Might they instead be derived merely
by reference to ‘objective’ indices (in the case of gender, that is, clothing,
behaviour, genitals, etc.; in the case of nationality, place of birth, passport
held, etc.)?
Central here is that a fully-fledged social identity requires recognition that
the self may be defined through relationship with other category members.
(Other features, too, are required; see Ruble et al., Chapter 2.) With this in
mind, we explored the initial development of the social self by means of two
92 Developmental aspects of social identity
studies. First, extending Bennett, Yuill, Banerjee, and Thomson’s (1998b)
work on identity in dyadic contexts, we examined children’s sense of
responsibility for actions committed by unknown in-group members. Second,
we explored how far, if at all, children evince a cognitive tendency to accentu-
ate self-stereotypical characteristics when relevant social identities are made
salient (as has been found among adults).

Study 6: Feelings of responsibility for in-group members’ actions

As Doosje, Branscombe, Spears, and Manstead (1998) have noted, an
important consequence of a social identity is that many aspects of self-
perception result not from one’s own actions, but from those of others who
share one’s social identity. For example, Doosje et al. note the sense of col-
lective guilt experienced by many Germans born following Word War II.
Since these individuals could in no way have personally influenced the events
of 1939–45, their sense of guilt and responsibility can be understood only in
terms of their social identity, that is, as Germans.
Previous research by Bennett et al. (1998b) has shown that not until around
7–8 years of age do children feel compromised by the wrongdoings of
another person with whom they are closely associated (mother or best
friend); prior to this they assert that others’ perceptions of them will be based
only on those actions for which they are personally responsible (e.g., “They’ll
still think I’m nice. He did it, it wasn’t me”). Although this study looked
only at interpersonal contexts, rather than the intergroup contexts required
to examine social identity, it nevertheless suggests the possibility that a
fully-fledged social identity may be a considerably later development than
previously assumed.
Using a method based on that of Bennett et al. (1998b), in collaboration
with Wendy Gibson, we investigated when children start to feel compromised
by the wrongdoings of in-group members. Participants (5-, 7-, and 9-year
olds) were asked to consider hypothetical scenarios in which either they per-
sonally were responsible for a particular negative outcome (personal condi-
tion), or in which it was stated that a member of the same social category
committed the action (social condition). For example, children were asked to
consider a context in which, during a sporting visit to another school, either
they or another (unknown) child from their school commit a normative
transgression. Following this, children were asked a variety of questions
about their likely responses to the events depicted. Interestingly, whereas in
the personal condition majorities in all age groups indicated that they would
want to apologise for the transgression, in the social condition the desire to
offer apology was present only in the 7- and 9-year-olds. Similarly, only
among the two older age groups was there a recognition of how the unknown
in-group member’s action would colour out-group members’ perceptions of
the in-group. These findings, like those of Bennett et al. (1998b), suggest that
the youngest children may have an essentially individualistic perspective:
 Sani and Bennett 93
That is, they may see actions as reflecting only upon their particular per-
petrators, rather than upon the groups of which those perpetrators are
members. Given this, one may question the widely held assumption that
young children’s references to social-categorical memberships carry the
same meaning as those of older children and adults.

Study 7: The emergence of self-stereotyping

As we noted earlier, a crucial issue to be addressed is when it is that children
become able to redraw the boundaries of the self in order to include other
people within the self. Study 5, conducted in collaboration with Georgina
Ferrier, investigated this issue in the context of gender identity. That is, we
studied when children begin to include the characteristics of their gender in-
group in the self. The inclusion of the in-group in the self was operationalised
as self-stereotyping, that is, as the perception of self in terms of the stereo-
typical characteristics of the gender in-group.
Our study included two experimental conditions. In both, the child was
asked to judge the self by specifying the extent to which a set of adjectives
(e.g., nice, friendly, brave, loud, clever, dirty, etc.) applied to him/her. More
precisely, the child had to say whether he/she was “not at all”, “quite”, or
“very” nice, friendly, brave, kind, and so on. However, in the first condition
the child had to describe only the self (“only self” condition), while in the
second condition he/she initially described the relevant gender out-group in
general – that is “girls’ if the child was a male, or “boys” if the child was a
female – and then subsequently the self (“out-group then self” condition).
Thus, this was a repeated measures design; the second condition was run 2
weeks after the first, to ensure that in the second condition participants would
not remember their previous judgments.
The central assumption of this study, then, was that if participants assign
more gender-stereotypical characteristics to themselves in the second than in
the first condition, then we can infer that the nature of the intergroup context
has made children’s social identity salient, and in turn, social identity has led
to important changes in self-perception and self-stereotyping. That is, we can
infer that the group has become part of the self.
Results show that 5-year-olds tended to describe themselves in a consistent
fashion across conditions: in-group identification does not seem to lead
younger children to a more gender-stereotypical perception of the self.
However, there is clear evidence that older males described the self in a
more gender-stereotypical fashion in the out-group then self than in the
“only self” condition. To start with, 7-year-old males defined self as bigger,
tougher, stronger, and more hardworking when judging it after judging
the group of girls than when judging it alone. Similarly, 10-year-old males
saw the self as dirtier and bigger in the out-group then self than in the “only
self” condition. Among girls, effects were less marked: As was the case for
5-year-olds boys, the youngest group of girls provided the same sorts of
94 Developmental aspects of social identity
self-descriptions regardless of condition. However, for both 7- and 10-year-
old females there was a marginal effect for the adjective “talkative”, in
the sense that they described the self as less talkative when gender identity
was salient. In sum, results indicate that by 7 years of age children become
cognitively able to include the in-group in the self. (The lack of clear
effects for girls may reflect the possibility that the adjectives used in the study
better reflected male than female identity, a possibility we are currently
addressing.)

Summary
Interestingly, the results of the two studies reported in this section, though
using quite different methodologies, seem to yield similar findings. Both sug-
gest that it is not until around 7 years of age, at the earliest, that children
internalise particular social identities. (And of course, the internalisation of
many identities, especially those based on belief, is likely to come rather later.)
Although we accept that much work is needed to substantiate this suggestion
properly, it points to the necessity to reconsider previous claims concerning
the early appearance of social identities. In particular, future research should
seek to distinguish between mere labelling of the self in social categorical
terms and internalisation of social categories in self-conception, in that it
may be that initial self-categorisation serves to direct attention to those
aspects of the environment that facilitate the subsequent internalisation of
an identity.
Despite convergence between the findings of the two studies, it may be
worth noting that the studies’ foci differ in at least one important way that
may transpire to be significant. The first study deals with phenomena that are
contingent upon having internalised a social identity whereas the second per-
haps deals more directly with presence or absence of an internalised identity.
Thus, in the former study we examined beliefs about out-group members’
views of the in-group, the desire to apologise, and so on. In the second study
we looked directly at self-conception. We speculate that more fine-grained
work may reveal the need to distinguish between the initial internalisation of
an identity and the subsequent coupling of internalisation to identity-related
processes (such as facework, social role-taking, social emotions, etc.). Such a
pattern of findings is well-documented in other domains. For example,
“theory of mind” research has shown that an understanding of false belief
precedes an appreciation of false belief-related emotions (e.g., feeling
happy because one falsely believes something positive has transpired; Harris,
Johnson, Hutton, Andrews, & Cooke, 1989). Thus, future research on social
identity should seek to distinguish measures of basic processes from those
contingent upon basic processes.
So far in our discussion we have spoken of the internalisation of social
identities as an all-or-none affair, and indeed, our measures, too, imply a
dichotomous conception in this respect. Inevitably, initial research on a
 Sani and Bennett 95
particular problem must make simplifying assumptions, and this is a case in
point. We fully accept, therefore, that future research should examine the
extent of identification with particular social categories. Indeed, not only may
identification vary quantitatively (i.e., in that children may feel more or less
identified with a group), but also qualitatively, in the sense that there are likely
to be age-related changes in category meaning (e.g., along the sorts of lines
implied by the first subsection of this chapter).

GENERAL SUMMARY AND CONCLUSIONS

The fundamental message of the three subsections of this chapter is that

social identity-related phenomena are subject to important developmental
changes. We have suggested that developments take place in conceptualising
identities. In particular, we propose that younger children conceive of iden-
tities largely in terms of particular behaviours, practices, and dispositions,
and that only later do they also consider socially shared beliefs. Second, we
suggest that children’s conceptualisation of identities is likely, with increasing
age, to vary with the comparative context in which conceptualisations are
instantiated. Finally, we propose that children’s internalisation of particular
social identities may be a somewhat later development than previous
researchers have supposed.
The work that we have outlined here represents only a start on these ques-
tions and much remains to be done in terms of detailed confirmation (or
not!) of our claims and their implications. We consider that the first two areas
(i.e., conceptualisations of identities and their contextual variability) are
fairly tractable methodologically and lend themselves readily to further
empirical inquiry. More problematic, in our view, is the assessment of the
extent to which social identities have penetrated children’s self-conception. In
looking at children’s subjective sense of themselves as group members we
would advocate (for the time being at least) the abandonment of conventional
methods, such as verbal self-description, on the grounds that categorical
labels can be applied with reference merely to public criteria, without neces-
sarily indicating subjective identification with a particular group. (See too
David et al., Chapter 5, who note also that social identification, though not
mere self-labelling, necessarily involves recognition of within-group similar-
ity and between-group difference.) Similarly, we have reservations about
methods that require children to rank-order identities, since these pre-
suppose that those identities are internalised (e.g., a child may rank-order
the identity “American” high not because it is one that is central to her
current sense of herself, but because this represents an identity aspiration,
one that others have marked out for her as desirable). In our own work on
this topic we have outlined two methods that we believe better assess the
internalisation of social identities. A key goal for future research is to extend
the range of methodologies that address this matter. We suggest that the
96 Developmental aspects of social identity
social psychological literature represents a rich resource in terms of
methodologies that might be adapted for developmental inquiry.
Like others in this volume, we propose that mid to late childhood may be a
particularly important time in terms of the understanding and internalisation
of many social identities, particularly those that are ascribed rather than
chosen (the latter perhaps being more characteristic of adolescence, e.g.,
Eckert, 1989). In addition to burgeoning cognitive developments that under-
lie categorical differentiation, it is clear that during this period children
acquire a social “face” (Bennett et al., 1998b; Miller, 1996) and recognise that
others make expectations in line with that face. Importantly, too, the shelter
afforded by the domestic arena of family life is to some extent diminished by
children’s participation in school life, and particularly the peer group. The
peer group, marked by greater heterogeneity than the family unit, inevitably
highlights issues of difference and increasingly calls attention to social
identities. Where social identities may have been relatively taken for granted
(or even invisible) within the family, they are likely to be subject to extensive
negotiation and even contestation within the peer group (see Connolly, 1998).
Such experiences may contribute not only to the development of the social
self but also to the personal self, inasmuch as, for example, children’s com-
mitments to public claims about the social self may sharpen their sense of
private selfhood.
In conclusion, we take the view that the heuristic value of the social iden-
tity approach to developmentalists is considerable, both in terms of empirical
predictions and conceptual illumination. The chapters of this book largely
attest to this contention. However, it is important to note that there are many
specifically developmental questions that are thrown up by this approach –
some of which we have tackled here, albeit relatively superficially, given that
our efforts have been largely descriptive rather than explanatory (that is, we
have said little of the possible causes underlying observed developments). It is
important, too, to reiterate the obvious but vital observation that these theor-
ies were developed with adult behaviour in mind. As such, they may reflect
assumptions that are untenable in a developmental context, thereby limiting
the theories’ developmental application3. Moreover, we suggest that there are
questions about the development of social identity that lie beyond the scope
of the social identity approach. For example, in considering the origins of
social identities, an essentially cognitive account, based on the individual
child’s capacity for categorisation of social stimuli, is likely to neglect

3 For example, social identity theory’s distinction between permeable and impermeable group
boundaries is predicated on an adult conception. From a developmental perspective it may be
a potentially problematic distinction in that, lacking ethnic/gender/and so on constancy, young
children are unlikely to respond to it in the same way, seeing all boundaries as in some sense
permeable. Similarly, concerning self-categorisation theory, the calculation of comparative
fit cannot be taken for granted in that the cognitive capacity to partition stimuli into sets and
subsets and reason about their interrelations is not apparent in young children.
 Sani and Bennett 97
important social sources of identity. As Jenkins (1996) has adeptly expressed
it, “if identity is a prerequisite for social life, the reverse is also true” (p. 20).
With this in mind, and drawing upon a Vygotskian analysis, we suggest
that it may be fruitful to conceive of the origins of social identities in terms
of co-construction (i.e., between novice and elder), rather than in terms of
purely individual cognitive construction. Drawing upon this approach,
one could then usefully think about the activities through which the social
enactment of identities, and the affirmation of those identities by others,
contributes to their internalisation. In turn, this raises the need to extend
the range of methodologies employed that have been used to date, to include,
for example, ethnographic and other qualitative techniques (Connolly,
2001).

Acknowledgements
The research reported in this chapter was largely funded by a grant from
the Economic and Social Research Council, UK (grant reference
R000222801).We are extremely grateful for the constructive comments made
by Martyn Barrett on an earlier version of this chapter.

REFERENCES

Abrams, D., & Hogg, M. (2001). Collective self. In M. A. Hogg & S. Tindale (Eds.),
Blackwell handbook of social psychology: Vol. 3. Group processes (pp. 425–461).
Oxford: Blackwell.
Banerjee, R., & Lintern, V. (2000). Boys will be boys: The effect of social evaluation
concerns on gender typing. Social Development, 9, 397–408.
Barrett, M., & Short, J. (1992). Images of European people in a group of 5–10-year-
old English schoolchildren. British Journal of Developmental Psychology, 10,
339–363.
Barrett, M., Wilson, H., & Lyons, E. (2003). The development of national ingroup
bias: English children’s attributions of characteristics to English, American and
German people. British Journal of Developmental Psychology, 21, 193–220.
Bar-Tal, D. (1998). Group beliefs as expressions of social identity. In S. Worchel,
J. Francisco Morales, D. Paez, & J.-C. Deschamps (Eds.), Social identity:
International perspectives (pp. 93–113). London: SAGE Publications.
Bartsch, K., & Wellman, H. M. (1995). Children talk about the mind. Oxford: Oxford
University Press.
Bennett, M., Lyons, E., Sani, F., & Barrett, M. (1998a). Children’s subjective
identification with the group and in-group favoritism. Developmental Psychology,
34, 902–909.
Bennett, M., Yuill, N., Banerjee, R., & Thomson, S. (1998b). Children’s understanding
of extended identity. Developmental Psychology, 34, 322–331.
Best, D. L., Williams, J. E., Cloud, M. J., Robertson, L. S., Edwards, J. R., Giles, H., &
Fowles, J. (1977). Development of sex-trait stereotypes among young children in the
United States, England and Ireland. Child Development, 48, 1375–1384.
98 Developmental aspects of social identity
Cairns, E. (1990). Impact of television news on children’s perceptions of violence in
Northern Ireland. Journal of Social Psychology, 130, 447–452.
Chi, M. T. H., & Rees, E. T. (1983). A learning framework for development. In
M. T. H. Chi (Ed.), Trends in memory development. Basel: Karger.
Connolly, P. (1998). Racism, gender identities and young children: Social relations in a
multi-ethnic inner-city primary school. New York: Routledge.
Connolly, P. (2001). Qualitative methods in the study of children’s racial attitudes and
identities. Infant and Child Development, 4, 219–233.
Damon, W., & Hart, D. (1988). Self-understanding in childhood and adolescence. New
York: Cambridge University Press.
Davey, A. (1983). Learning to be prejudiced. London: Edward Arnold.
Deaux, K., Reid, A., Mizrahi, K., & Cotting, D. (1999). Connecting the person to
the social: The functions of social identification. In T. R. Tyler, R. M. Kramer, &
O. P. John (Eds.), The psychology of the social self (pp. 91–114). Mahwah, NJ:
Lawrence Erlbaum Associates, Inc.
Doosje, B., Branscombe, N. R., Spears, R., & Manstead, A. S. R. (1998). Guilty by
association: When one’s group has a negative history. Journal of Personality and
Social Psychology, 75, 872–886.
Eckert, P. (1989). Jocks and burnouts: Social categories and identity in the high school.
Columbia, SC: Teachers College Press.
Eckes, T., & Trautner, H. M. (Eds.) (2000). The developmental social psychology of
gender. Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Eder, R. A. (1990). Uncovering young children’s psychological selves: Individual and
developmental differences. Child Development, 61, 849–863.
Edwards, C. P. (1984). The age group labels and categories of preschool children.
Child Development, 55, 440–452.
Fagot, B. L., & Hagan, R. (1991). Observations of parents reactions to
sex-stereotyped behaviors. Child Development, 62, 1617–1628.
Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impression formation, from
category-based to individuating processes: Influences of information and
motivation on attention and interpretation. In M. P. Zanna (Ed.), Advances in
experimental social psychology, Vol. 23 (pp. 1–74). New York: Academic Press.
Hamilton, D. L., & Trolier, T. K. (1986). Stereotypes and stereotyping: An overview
of the cognitive approach. In J. F. Dovidio & S. Gaertner (Eds.), Prejudice,
discrimination and racism (pp. 127–163). London: Academic Press.
Harris, P., Johnson, C. N., Hutton, D., Andrews, G., & Cooke, T. (1989). Young
children’s theory of mind and emotions. Cognition and Emotion, 3, 379–400.
Harter, S. (1999). The construction of the self: A developmental perspective. New York:
Guilford Press.
Haslam, S. A., & Turner, J. C. (1992). Context-dependent variation in social stereo-
typing 2: The relationship between frame of reference, self-categorization and
accentuation. European Journal of Social Psychology, 22, 251–278.
Haslam, S. A., Turner, J. C., Oakes, P. J., McGarty, C., & Hayes, B. K. (1992). Context-
dependent variation in social stereotyping 1: The effects of intergroup relations
as mediated by social change and frame reference. European Journal of Social
Psychology, 22, 3–20.
Hogg, M. (2001). Social identity and social comparison. In J. Suls & L. Wheeler
(Eds.), Handbook of social comparison (pp. 401–421). Hingham, MA: Kluwer
Academic Publishers.
 Sani and Bennett 99
Hopkins, N., & Murdoch, N. (1999). The role of the “other” in national identity:
Exploring the context-dependence of the national in-group stereotype. Journal of
Community and Applied Social Psychology, 9, 321–338.
Hopkins, N., Regan, M., & Abell, J. (1997). On the context dependence of
national stereotypes: Some Scottish data. British Journal of Social Psychology,
36, 553–563.
Jenkins, R. (1996). Social identity. London: Routledge.
Lambert, W. E., & Klineberg, O. (1967). Children’s views of foreign peoples. New
York: Appleton-Century-Crofts.
Martin, C. L., & Halverson, C. F. (1981). A schematice processing model of sex typing
and stereoptyping in children. Child Development, 52, 1119–1134.
Miller, R. E. (1996). Embarrassment: Poise and peril in everyday life. New York:
Guilford Press.
Oakes, P. J., Haslam, A., & Turner, J. C. (1994). Stereotyping and social reality. Oxford:
Blackwell.
Perner, J. (1999). Theory of mind. In M. Bennett (Ed.), Developmental psychology:
Achievements and prospects. New York: Psychology Press.
Pomerantz, E. M., Ruble, D., Frey, K. S., & Greulich, F. (1995). Meeting goals and
confronting conflict: Children’s changing perceptions of social comparison. Child
Development, 66, 723–738.
Powlishta, K. K. (2002). Perceived similarity of boys and girls: Contextual variations in
the salience of gender. Manuscript under review.
Premack, D., & Premack, A. J. (1995). Origins of human social competence. In
M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 205–218). Cambridge, MA:
MIT Press.
Quintana, S. M. (1998). Children’s developmental understanding of ethnicity and
race. Applied and Preventive Psychology, 7, 27–45.
Ruble, D. N. (1983). The development of social comparison processes and their role in
achievement-related self-socialization. In E. T. Higgins, D. N. Ruble, & W. W.
Hartup (Eds.), Social cognition and social development: A socio-cultural perspective.
Cambridge: Cambridge University Press.
Ruble, D. N., & Flett, G. L. (1988). Conflicting goals in self-evaluative information
seeking. Child Development, 59, 97–106.
Ruble, D., & Frey, K. (1991). Changing patterns of comparative behavior as skills
are acquired. In J. Suls & T. A. Wills (Eds.), Social comparison: Contemporary
theory and research (pp. 70–112). Hillsadale, NJ: Lawrence Erlbaum Associates,
Inc.
Sani, F., & Bennett, M. (2001). Contextual variability in young children’s gender
ingroup stereotype. Social Development, 10, 221–229.
Sani, F., Bennett, M., Agostini, L., Malucchi, L., & Ferguson, N. (2000). Children’s
conception of characteristic features of category members. Journal of Social
Psychology, 140, 227–239.
Sani, F., Bennett, M., & Joyner, L. (1999, 13–15 September). Developmental aspects of
contextual variability in the stereotype of the ingroup: The case of Scottish children.
Paper presented at the Annual Conference of the Social Psychology Section of the
British Psychological Society, Lancaster.
Sani, F., Bennett, M., Mullally, S., & MacPherson, J. (2003). On the assumption of
fixity in children’s stereotypes: A reappraisal. British Journal of Developmental
Psychology, 21, 113–124.
100 Developmental aspects of social identity
Sani, F., Lyons, E., & Barrett, M. (1996). The child’s discovery of the collective self.
Paper presented at the Annual Conference of the Social Section of the British
Psychological Society, University of Oxford.
Sani, F., & Thomson, L. (2001). We are what we wear: The emergence of consensus
in stereotypes of students and managers’ dressing style. Social Behaviour and
Personality, 29, 695–700.
Searle, J. (1983). Intentionality: An essay in the philosophy of mind. Cambridge:
Cambridge University Press.
Serbin, L. A., & Sprafkin, C. (1986). The salience of gender and the process of
sex-typing in three- to seven-year-old children. Child Development, 57, 1188–1199.
Stangor, C. (1995). Stereotyping. In A. S. R. Manstead & M. Hewstone (Eds.), The
Blackwell encyclopedia of social psychology (pp. 628–633). Oxford: Blackwell.
Stangor, C., & Shaller, M. (1996). Stereotypes as individual and collective repre-
sentations. In C. N. Macrae, C. Stangor, & M. Hewstone (Eds.), Stereotypes and
stereotyping (pp. 3–37). New York: Guilford Press.
Tajfel, H. (1981). Human groups and social categories. Cambridge: Cambridge
University Press.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. (1987).
Rediscovering the social group: A self-categorisation theory. Oxford: Blackwell.
Turner, J. C., & Oakes, P. J. (1997). The socially structured mind. In C. McGarty &
S. A. Haslam (Eds.), The message of social psychology (pp. 355–373). Oxford:
Blackwell.
Turner, J. C., & Onorato, R. S. (1999). Social identity, personality, and the self-
concept: A self-categorization perspective. In T. R. Tyler, R. M. Kramer, & O. P.
John (Eds.), The psychology of the social self (pp. 11–46). Mahwah, NJ: Lawrence
Erlbaum Associates, Inc.
Turner, J. C., Oakes, P. J., Haslam, S. A., & McGarty, C. (1994). Self and collective:
Cognition and social context. Personality and Social Psychology Bulletin, 20,
454–463.
Tyler, T. R., Kramer, R. M., & John, O. P. (Eds.) (1999). The psychology of the social
self (pp. 11–46). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Part II

Identities
4 Gender as a social category:
Intergroup processes and
gender-role development
Kimberly K. Powlishta

“All boys push! All boys push!” is a chant I recall from my childhood, as a
group of girls on the playground during recess at school tried to convince
those “other creatures” to push the merry-go-round. Boys too seem to
view the sexes as members of distinct groups, as illustrated in the following
lunchroom scene:

a fourth-grade boy pointed at one of the white squares that alternated

with green squares on the linoleum floor. “That’s kook territory . . . girl’s
territory,” he said loudly, tiptoeing in an exaggerated way from one green
square to another. “If you step on the white you change into a girl.” He
tiptoed around a girl and then leaned up against the wall. “This is boys’
territory by the wall,” he said.
(Thorne, 1987, p. 10)

With this sense that boys and girls are very different from each other,
having separate responsibilities and territories, often comes the belief that
one’s own sex is better. The other sex “has cooties” and should be avoided.
For example, the boy quoted above equated girls with “kooks”. Similarly, an
11-year-old girl told Maccoby and Jacklin (1987) that “Nobody who had any
care of status would sit next to a boy if they could sit next to a girl . . . It is
sort of like being in a lower rank or peeing in your pants” (p. 245).
More formal research evidence supports these observations that boys and
girls tend to exaggerate differences between the sexes and to show strong
biases favouring their own sex. The current chapter will review this evidence
and propose that generic intergroup processes (i.e., in-group favouritism, the
accentuation of within-group similarities and between-group differences,
social stereotyping, and out-group homogenisation) contribute to children’s
beliefs about, attitudes towards, and interactions with, other males and
females. The extent to which developmental factors, individual and group
differences, and variations in social context may influence the salience of
gender as a social category and the subsequent activation of these intergroup
processes also will be discussed. Finally, the ways in which gender may differ
from other social categories will be described.
104 Gender as a social category
INTERGROUP PROCESSES

Numerous social psychological studies, conducted primarily with adolescents

and adults, have demonstrated that mere assignment to a social category can
lead people to favour members of their own groups. In the original studies by
Tajfel and colleagues (e.g., Tajfel, 1970; Tajfel, Billig, Bundy, & Flament,
1971), participants were assigned at random to one of two groups defined on
some apparently trivial basis (e.g., did the person over- or underestimate the
number of dots in an array?). Even though group membership is anonymous,
and group members have no conflicts of interest, pre-existing hostilities, or
face-to-face contacts, individuals placed in these sorts of “minimal groups”
show favouritism towards in-group members (e.g., a fellow “dot underesti-
mator”). They display this bias when allocating rewards, making ratings
on positive and negative traits, or evaluating group products (Brown, 2000;
Messick & Mackie, 1989). In other words, as a mere result of being categor-
ised, evaluative perceptions are distorted. According to social identity theory
(Tajfel & Turner, 1979), individuals seek to differentiate their own groups
positively from others in this way in order to enhance their own self-esteem
(i.e., to achieve a positive social identity).
Although studied less frequently, children also display in-group favourit-
ism in laboratory minimal group situations. In one of the earliest studies,
7- to 11-year-olds divided randomly into groups, supposedly on the basis of
the type of art they preferred, allocated more money to in-group than to
out-group others. In fact, they showed as much favouritism as when asked to
divide the rewards between their “best friend” and someone they did not
like (Vaughan, Tajfel, & Williams, 1981). Similar minimal group in-group
favouritism effects have been obtained with European, Maori, and Samoan
8-year-olds in New Zealand (Wetherell, 1982) and with 10- to 12-year-old
English boys (Moghaddam & Stringer, 1986).
More recently, Bigler and colleagues (Bigler, 1995; Bigler, Jones, &
Lobliner, 1997) found that 6- to 11-year-old children showed favouritism
towards members of their own, randomly assigned “colour groups” (e.g.,
red vs. green), at least if the group distinction was made salient through the
functional use of groupings by teachers in the classroom. They displayed this
favouritism in a variety of ways, for example by being unwilling to change
groups, by predicting that their own group would win contests, by choosing
more in-group than out-group members to go on a field trip, and by estimat-
ing that in-group children had received more rewards for good behaviour
than had out-group children. In addition, when the grouping distinction
was made perceptually salient through the use of coloured tee shirts (Bigler
et al., 1997), children reported liking individual in-group classmates more
than out-group classmates. Relative to a control condition, children whose
teachers made functional use of the colour groupings also showed more
in-group favouritism in their positive and negative trait attributions.
Using a minimal group paradigm in which children were assigned at
 Powlishta 105
random to higher (fast) or lower (slow) status teams, Yee and Brown (1992)
also demonstrated in-group favouritism. Children reported liking their own
team more than the other as early as 3 years of age, regardless of status. Five-
year-olds were particularly biased, rating their own team as faster than the
other even when assigned to the “slow” team. Nesdale and Flesser (2001)
demonstrated consistent in-group favouritism in the liking ratings of children
in a minimal group paradigm as well. The extent of this favouritism varied,
however, depending on the relative status of the in-group and out-group, the
permeability of group boundaries, and the availability of non-status-related
dimensions on which to compare the groups, consistent with earlier studies
of adults.
In addition to in-group favouritism, social categorisation also leads to dis-
torted perceptions of variability (Brown, 2000; Messick & Mackie, 1989). In
particular, people tend to exaggerate differences between and similarities
within groups along non-evaluative as well as evaluative dimensions. In fact,
it has been argued that these perceptual distortions (i.e., cognitive differen-
tiation) may contribute to in-group favouritism, above and beyond the need
for self-esteem maintenance posited by social identity theory (Doise &
Sinclaire, 1973; Wilder, 1981). In other words, perceptual accentuation of
between-group differences and within-group similarities, when coupled with
positive views of the self, would lead to in-group favouritism. This perceptual
accentuation effect also may encourage the formation and use of social
stereotypes, as stereotypes are based on the notion that members of a particu-
lar social category are alike in some way, a way that distinguishes them from
other groups (Brown, 2000). The exaggeration of within-group similarities is
often especially pronounced when evaluating groups to which one does not
belong (“they’re all alike, but we’re individuals”). That is, out-groups are
believed to be more homogeneous than in-groups (Quattrone, 1986), at least
on some dimensions (Brown, 2000).
As with in-group favouritism, although most of the research concerning
the impact of mere social categorisation on perceptions of variability has
been conducted with adults, there has been some research with children.
Bigler et al. (1997) reported that, relative to a control group, children whose
teachers made functional use of experimentally created groups perceived
greater between-group differences. These children also made more extreme
judgments concerning the in-group than the out-group. However, because the
judgments all involved evaluative trait ratings, the apparent accentuation of
between-group differences (in-group ratings higher on positive traits and
lower on negative traits than out-group ratings) and homogenisation of the
in-group (“all of us have positive traits; none of us have negative traits”) are
confounded with in-group favouritism. Nesdale and Flesser (2001) measured
perceptions of similarity in a minimal group situation more directly, finding
that in general, 5- and 8-year-olds rated themselves as more similar to the
in-group than to the out-group.
In summary, even in the absence of any pre-existing attitudes based on
106 Gender as a social category
direct experience or environmental teachings, the mere act of placing people
into categories can set in motion intergroup processes whereby the in-group is
favoured, differences between and similarities within groups are exaggerated,
and the out-group is often homogenised.
These same sorts of phenomena also can be seen with naturally occurring
groups (Brown, 2000; Messick & Mackie, 1989). Children begin to notice
such real-world social categories at a very early age. Cognitively grouping
people allows for the efficient processing of large amounts of information,
while at the same time helping the child to establish a self-identity (Martin &
Halverson, 1981; Serbin, Powlishta, & Gulko, 1993). During childhood, one
of the most salient groupings is based on gender (Powlishta, 2002b; Serbin et
al., 1993). There are a number of reasons why this may be so. First, people in
general (Allport, 1954), but especially young children (Livesley & Bromley,
1973), tend to focus on perceptually salient characteristics. As Allport (1954)
put it nearly 50 years ago:

Visibility and identifiability aid categorization . . . All our experience

teaches us that when things look different they usually are different. A
black cloud in the sky has very different significance from a white cloud.
A skunk is not a cat. Our comfort and sometimes our lives depend on
learning to act differently in the face of unlike objects.
(pp. 129, 131)

Like black and white clouds, the distinction between males and females is
perceptually salient. Actual differences between the sexes make gender group-
ings useful for predicting and monitoring behaviour. By definition, males and
females play different roles in reproduction. Sex also is a stable, dichotomous,
exhaustive, biological, “natural kind” basis for categorisation. Furthermore,
the gender distinction is emphasised by both adults and peers in the child’s
environment. Hence, it is not surprising that sex-based categorisation is so
prevalent during childhood (Bigler al., 1997; Martin & Halverson, 1981;
Serbin et al., 1993), typically occurring at a much earlier age than other more
abstract forms of social categorisation, such as those based on nationality
(Rutland, 1999).
If children are treating boys and girls as members of distinct groups, then
the same sort of generic intergroup processes demonstrated in minimal group
studies might influence children’s perceptions, beliefs, attitudes, and
behaviours towards the sexes as well. In other words, viewing males and
females as “us” versus “them” may lead children to favour their own sex,
exaggerate differences between and similarities within each sex, hold stereo-
typic beliefs and expectations, and view the other sex as more homogeneous
than their own. In this way, generic intergroup processes may contribute to
gender-role development.
 Powlishta 107
THE DEVELOPMENT OF GENDER CATEGORIES

The first prerequisite for intergroup processes to contribute to gender-role

development is that children must be able to categorise themselves and others
into male and female groups. The groundwork for this categorisation is laid in
infancy. A number of studies have demonstrated that infants will lose interest
in a particular novel face or voice that is presented repeatedly (habituation).
When the face or voice is changed, interest is regained (dishabituation), but
only if the sex of the new target is different from the original. In other words,
even dissimilar targets fail to elicit distinctive reactions if they are of the same
sex. However, a male and a female are treated as different by the middle of the
first year, suggesting the existence of rudimentary gender categories among
these infants (Fagan & Shepherd, 1981; Fagan & Singer, 1979; Katz, 1996;
Leinbach & Fagot, 1993; Miller, 1983).
Not long after showing the ability to recognise gender categories in this
way, children more actively begin to divide the world into males and females.
For example, Johnston, Madole, Bittinger, and Smith (2001) gave toddlers a
set of male and female dolls and examined the sequence in which the dolls
were touched. Touching dolls from a single gender category in succession
more than would be expected by chance was used to infer gender categorisa-
tion. Such categorisation showed a sharp increase between 18 and 22 months
of age.
These gender categories soon become even more explicit. Children are able
to sort photographs on the basis of gender by 2 to 3 years of age (Katz, 1996;
Sen & Bauer, 2002; Thompson, 1975; Weinraub, Clemens, Sockloff, Ethridge,
Gracely, & Meyers, 1984; Yee & Brown, 1994). Two-year-olds often begin to
use gender labels appropriately as well, becoming experts at labelling them-
selves and others by the time they are three years old (Katz, 1996; Leinbach &
Fagot, 1986; Levy, 1999; Sen & Bauer, 2002; Thompson, 1975; Weinraub
et al., 1984).
Gender appears to be quite salient at this age. For example, Serbin and
Sprafkin (1986) presented young children with a target male or female (e.g., a
photo of a man stirring a pot) and asked them to select another photo that
“goes with” the first from among three options: one that matched the target
in terms of gender (e.g., a man reading), one that matched on some other
dimension such as activity, body stance, or facial expression (e.g., a woman
rolling dough), and a third that did not match on any obvious dimension (e.g.,
a woman sweeping). Three-year-olds tended to make gender-based matches.
Although children became more likely to focus on alternative dimensions
rather than gender as they got older in this study (Serbin & Sprafkin, 1986),
other methodologies (and everyday experience) suggest that gender remains a
salient basis for categorisation throughout childhood. For example, two stud-
ies using a release from a proactive interference paradigm have demonstrated
that elementary-school-aged children encode the gender connotations of
words in memory. That is, when presented either with all masculine or all
108 Gender as a social category
feminine words to remember on a series of trials, children (especially girls)
show an increase in recall on a subsequent trial in which the words are drawn
from the other gender category (Kail & Levine, 1976; Perez & Kee, 2000).
When it comes to categorising people (as opposed to words), both sexes
show evidence of spontaneously invoking gender in their automatic pro-
cessing of social information during mid and late childhood. Bennett and
colleagues (Bennett, Sani, Hopkins, Agostini, & Malucchi, 2000) presented
children with a series of statements attributed to individual boys and girls
depicted in photographs. When asked to recall “who said what?”, children
were more likely to be confused about which boy or girl made a particular
statement than about whether the statement was made by a boy or a girl. The
children thus seemed to be categorising the people and their statements in
terms of gender.
Additional evidence that children of this age spontaneously form gender-
based person categories comes from a study in which 6- to 12-year-olds were
presented with photographs of people and asked to rate the similarity of each
pair (Powlishta, 2002b). A multidimensional scaling procedure was used to
obtain a spatial configuration of the photographs, with pairs receiving higher
similarity ratings placed closer together in space. The features presumably
influencing similarity judgments are thus represented as dimensions in the
resulting configuration. Four such dimensions were revealed – gender, age,
facial expression, and hair colour – with gender being the most salient. Once
again, then, there is evidence that gender is ubiquitous as a social category
(Maccoby, 1988).

EXAGGERATION OF BETWEEN-SEX DIFFERENCES AND

WITHIN-SEX SIMILARITIES

Given that children do seem to treat gender as an important basis for categor-
ising their social world, is there evidence that the same sorts of intergroup
processes seen in minimal group studies of children and adults also influence
children’s perceptions, attitudes, and behaviours regarding males and
females? Children do often treat the sexes as very different from each other,
while emphasising the ways in which all boys or all girls are similar. For
example, elementary-school-aged children rate unfamiliar people as more
similar to themselves (Brewer, Ho, Lee, & Miller, 1987; Powlishta, 1995c) or
two unfamiliar people as more similar to each other (Powlishta, 2002b) when
they are of the same sex.
Children also tend to assume that same-sex peers will have similar interests
to their own. For example, Martin, Eisenbud, and Rose (1995) showed pre-
schoolers unfamiliar, non-sex-typed toys and asked them to rate how much
they and other children would like each toy. Despite the fact that boys and
girls showed a similar liking of the toys in their self-ratings (confirming that
the toys were not traditionally sex-typed), their self and other ratings were
 Powlishta 109
more similar when the others were of their own sex. In other words, children
made the gender-based inference that “What I like, children of my sex will
also like, and children of the other sex will not like” (p. 1453).
More generally, children seem quite ready to form new gender-based
stereotypes. In fact, this tendency may even begin in infancy. When presented
with a series of male or female faces paired with objects, 10-month-olds
increase their attention to a new face–object pairing only when a face of one
sex is paired with an object previously associated with the other sex. These
findings suggest that infants can detect correlations between gender and
other attributes, in a sense forming primitive gender stereotypes (Levy &
Haaf, 1994).
This readiness to attach new characteristics to gender becomes even more
apparent in the preschool years. Gelman, Collman, and Maccoby (1986)
taught 4-year-olds new sex-linked properties (e.g., “This boy has little seeds
inside; this girl has little eggs inside”). When shown another labelled boy or
girl, children were able to infer sex-linked properties (e.g., that a new “boy”
would also have seeds inside). Similarly, Bauer and Coyne (1997) assigned a
different, traditionally gender-neutral preference to a boy and a girl figure.
When asked about the preference of a new ambiguous-looking target labelled
as either a boy or girl, 3½-year-olds made stereotypical inferences (i.e., that
the “boy” target would prefer an object similar to the one liked by the ori-
ginal boy and the “girl” target would prefer an object similar to the one liked
by the original girl).
In addition to readily attaching new characteristics to gender categories,
children also have extensive knowledge of traditional gender stereotypes.
This knowledge can be seen as early as the toddler years. Using a visual
preference paradigm in a series of studies, Serbin, Poulin-Dubois, and col-
leagues have demonstrated that 18- to 24-month-olds have at least some
awareness of toy, activity, and metaphorical (e.g., bears are masculine) gender
associations (Eichstedt, Serbin, Poulin-Dubois, & Sen, in press; Poulin-
Dubois, Serbin, Eichstedt, & Sen, in press; Serbin, Poulin-Dubois, Colburne,
Sen, & Eichstedt, 2001). Knowledge of activity stereotypes is also seen in 2-
year-olds when they choose a “sex-appropriate” doll to imitate gender-
stereotyped actions (e.g., shaving the face, putting on lipstick; Poulin-Dubois
et al., in press).
When children are asked more directly whether certain toys, objects,
activities, occupations, or clothing are associated with males or females,
their stereotype knowledge exceeds chance levels by 2 to 2½ years of age
and increases rapidly throughout the preschool (Blaske, 1984; Edelbrock &
Sugawara, 1978; Helwig, 1998; Katz, 1996; Kuhn, Nash, & Brucken, 1978;
Leinbach, Hort, & Fagot, 1997; Martin & Little, 1990; Ruble & Martin,
1998; Thompson, 1975; Vener & Snyder, 1966; Weinraub et al., 1984) and
elementary-school-aged years (Carter & Patterson, 1982; Nadleman, 1974;
Ruble & Martin, 1998; Trautner, Helbing, Sahm, & Lohaus, 1989; Vener &
Snyder, 1966). In fact, knowledge of these sorts of stereotypes reaches ceiling
110 Gender as a social category
on many measures during early childhood (Serbin et al., 1993; Signorella,
1987).
Awareness of the gender stereotyping of personality traits lags behind the
more concrete object and activity stereotypes, but even 2½- to 3-year-olds
show at least minimal awareness of them (Albert & Porter, 1983; Cowan &
Hoffman, 1986; Etaugh & Riley, 1979; Haugh, Hoffman, & Cowan, 1980;
Reis & Wright, 1982). Substantial increases in personality stereotype know-
ledge occur between the ages of 5 and 11 years, so that adult-like levels of
knowledge are shown by late childhood (Beere, 1990; Best, 1982; Best et al.,
1977; Serbin et al., 1993; Williams, Bennett, & Best, 1975).
Children’s knowledge and acceptance of gender stereotypes also can be
seen in the sorts of inferences they make about other people. Even preschoolers
can use information about the stereotyped preferences and characteristics of
a gender-unspecified target to predict that the target would like other toys
and clothing consistent with that stereotype (e.g., someone who likes a toy
soldier would probably also like cars; someone who is a librarian would
probably wear a two-piece bathing suit rather than swimming trunks; Bauer,
Liebl, & Stennes, 1998; Martin, Wood, & Little, 1990). Older children make
increasingly complex inferences, using a person’s gender or gender-typed
characteristics to make predictions about competencies, behaviours, toy pref-
erences, personality traits, physical characteristics, roles, and occupations
(Berndt & Heller, 1986; Biernat, 1991; Cann & Palmer, 1986; Martin, 1989;
Martin et al., 1990; Powlishta, 1995c, 2000).
The fact that children are so knowledgeable about and willing to make use
of traditional gender stereotypes suggests that they may be showing the same
sort of exaggeration of between-group differences and within-group similar-
ities seen in minimal group experiments. On the other hand, it is possible that
children are reporting actual differences and similarities involving males and
females, rather than exaggerating them.
On average, boys and girls certainly do differ in many ways. They tend to be
more similar than they are different, however (see Powlishta, Sen, Serbin,
Poulin-Dubois, & Eichstedt, 2001, and Ruble & Martin, 1998, for recent
reviews). As Maccoby (1998) points out, most existing sex differences reflect
highly overlapping distributions, with extensive within-sex variability in the
sorts of preferences, interests, aspirations, traits, skills, and interpersonal
styles children display. Only a few behaviours are highly differentiated by sex
(i.e., playmate preferences, rough-and-tumble play, direct aggression, and the
themes enacted in pretend play). Similarly, Ruble and Martin (1998) note that
on average boys are more aggressive, assertive, and active, less socially
oriented and sensitive, better at physical activities, and have higher self-
esteem than girls. But they found little evidence to support other gender
stereotypes, such as those involving prosocial behaviour, moral reasoning,
passivity, or dependence.
Nevertheless, young children often treat gender stereotypes as if they are
rigidly binding absolutes. We’ve seen that they are quite willing to learn or
 Powlishta 111
create new stereotypes (e.g., that boys have little seeds inside), stereotypes that
they are unlikely to have encountered in their everyday world. Carol Martin
(2000) reports an anecdote in which her 4-year-old niece even created a
stereotype (girls but not boys have eyelashes) that obviously contradicts
reality. My 3½-year-old nephew similarly informed me recently that he likes
“boy songs” but not “girl songs”. He was unable to describe the difference,
and when I asked him to define “boy songs”, all he could tell me is that
“they’re not girl songs”. He also let me know that because he’s a boy I should
refer to him as “handsome” rather than “cute”, apparently generalising from
the more common handsome/pretty distinction.
Once they are formed, children’s gender stereotypes tend to be highly
resistant to change (Katz, 1986). The fact that children show better memory
for information that is consistent rather than inconsistent with their stereo-
types, even distorting information so that it becomes stereotype-consistent
(e.g., relabelling a male nurse as a doctor), may contribute to this resistance
(Cordua, McGraw, & Drabman, 1979; Martin & Halverson, 1983; Stangor &
Ruble, 1987). Children also tend to evaluate negatively anyone who violates
stereotyped norms, particularly in early to mid childhood (Damon, 1977;
Levy, Taylor, & Gelman, 1995; Martin, 1989). This ready acceptance of new
gender stereotypes and resistance to stereotypic changes or violations sug-
gests that children’s stereotypes represent more than a truthful reflection of
their world. Instead, children do seem to be exaggerating similarities within
and differences between the sexes.
Very little research has examined whether children perceive the other sex to
be more homogeneous than their own. With adults, anecdotal evidence sug-
gests that other-sex homogenisation sometimes occurs. One often hears
people making generalised statements about the other sex (e.g., a man saying
“I don’t understand women”; a woman saying that “Men are so—”) but
rarely about their own sex. Research evidence backs up this observation. In
a series of studies, Park and Rothbart (1982) found that stereotypic charac-
teristics were viewed as more prevalent in a given sex when members of the
other sex were doing the rating. Counterstereotypic trait ratings showed the
reverse pattern. These findings indicate that people have a more complex and
varied image of their own gender group than they do of the other.
The extent to which this phenomenon is seen in children is not entirely
clear. One study has provided some tentative evidence for other-sex hom-
ogenisation in childhood. In this study, children viewed unfamiliar boys and
girls in videotaped segments and rated the extent to which they thought
each target possessed a series of traits. A measure reflecting the variability
in ratings given to the three targets of each sex was created. For male
targets only, this measure was higher among boys than among girls. That is,
there was some evidence for other-sex homogenisation by girls (Powlishta,
1995c).
112 Gender as a social category
OWN-SEX FAVOURITISM

In addition to exaggerating between-sex differences and within-sex similar-

ities, extensive evidence indicates that children show own-sex favouritism in
their behaviours, attitudes, and perceptions (see Glick & Hilt, 2000, for a
recent review of “gender prejudice”). For example, even 2½- to 3-year-olds
have begun to prefer same-sex playmates. This gender segregation becomes
more prevalent with age, at least through early or mid childhood (see
Maccoby, 1998, for a review). In fact, in one study children were observed
with same-sex playmates 11 times as often as with other-sex playmates
(Maccoby & Jacklin, 1987). When asked to name their friends, older children
rarely list someone of the other sex (Hayden-Thomson, Rubin, & Hymel,
1987; Powlishta, 2001; Powlishta, Serbin, Doyle, & White, 1994; Serbin et al.,
1993). They also report liking same-sex classmates more than other-sex
classmates (Powlishta & Vartanian, 1999). And in one study, both boys and
girls estimated that classmates of their own sex had performed better on a
series of games than had classmates of the other sex (Deschamps & Doise,
1978). In other words, the preference for same-sex peers is so pervasive that it
is found seemingly no matter what the measure. When I have carried out
research with groups of children and have looked back as they are following
me from the classroom to the testing room, they typically are walking in
a gender-segregated fashion. That is, I see anecdotal evidence for own-sex
favouritism even before the research has officially begun.
At least in the elementary-school-aged years, this preference for one’s own
sex extends beyond children’s attitudes and behaviours toward familiar peers.
Children also show biases towards male and female strangers. For example,
when asked to rate their liking for boys and girls depicted in pictures, video-
tapes, or short verbal descriptions, children give higher ratings to targets of
their own sex (Martin, 1989; Powlishta, 1995c, 2001; Powlishta et al., 1994).
They also predict that such targets will have more positive traits and fewer
negative traits than targets of the other sex (Powlishta, 1995c, 2001; Zalk &
Katz, 1978). When shown drawings depicting unfamiliar boy–girl or man–
woman pairs and asked to choose who they would like to play with, children
again show a significant bias in favour of their own sex (Powlishta et al., 1994;
Serbin & Sprafkin, 1986).
In addition to favouring both familiar and unfamiliar same-sex individuals,
children of this age also believe that their own sex is better in general. This
effect can be seen when children are asked to provide global ratings of how
they feel about boys and girls (Yee & Brown, 1994). Children also show
earlier knowledge of personality trait stereotypes that portray their own sex
in a favourable light (Serbin et al., 1993). Furthermore, they attribute more
positive and fewer negative characteristics to their own sex than to the other
(Deschamps & Doise, 1978; Parish & Bryant, 1978; Powlishta, 1995a;
Powlishta et al., 1994; Powlishta & Vartanian, 1999; Silvern, 1977; Yee &
Brown, 1994).
 Powlishta 113
Of course, it is possible that children’s own-sex favouritism does not reflect
biased perceptions, per se. Instead, as with gender stereotypes, such prefer-
ences may result from actual differences between boys and girls. For example,
children may seek out same-sex playmates not because they are biased against
the other sex, but instead because sex differences in the interests or play styles
of boys and girls make same-sex playmates more compatible (Maccoby, 1998;
Powlishta, 1995b; Serbin, Moller, Gulko, Powlishta, & Colburne, 1994). Such
compatibility may lead them to have more friendships with and greater liking
for familiar same-sex peers. Rational inferences based on these past experi-
ences could then lead children to predict that they would like same-sex
strangers more than other-sex strangers and to develop more positive general
attitudes about their own sex.
Although actual experiences with boys and girls undoubtedly do contrib-
ute to children’s own-sex preferences (Maccoby, 1998), there is good evidence
that perceptions are also distorted by in-group favouritism. For example,
boys and girls generally agree on what traits are positive or negative. Yet the
very same trait is rated as more masculine/less feminine by boys than by girls
if it is positive (e.g., strong; gentle), and by girls than by boys if it is negative
(e.g., messy; cries; Powlishta, 1995a; Silvern, 1977). In other words, “children
do not simply like what they see in members of their own sex; they see what
they like” (Powlishta et al., 2001, p. 124).

THE IMPACT OF SOCIAL CONTEXT ON THE SALIENCE

OF GENDER

According to the more recent elaboration of social identity theory known as

self-categorisation theory (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987;
Turner & Onorato, 1999), the way in which the self is categorised depends on
the social context. That is, some contexts heighten the salience of personal
identity, wherein the self is defined in terms of personal or idiosyncratic attrib-
utes. Other contexts heighten the salience of social identity, wherein the self is
defined in terms of membership in social categories. Intergroup processes (i.e.,
in-group favouritism, exaggeration of intragroup similarities and intergroup
differences, social stereotyping) should be most apparent in the latter contexts.
Which social categories are salient can also vary with context. “People who
are categorized and perceived as different in one context (e.g., biologists and
physicists in a science faculty) can be recategorized and perceived as similar in
another contest (e.g., scientists rather than social scientists in a university)
without any actual change in their own positions” (Turner & Onorato, 1999,
p. 23). Or as Taylor (1981) put it:

An apple is less likely to be categorized as a fruit if it is with the objects:

beach ball, cube, cardboard box, and ball bearing than if it is with the
objects: orange, carrot, bean, and pear. Likewise, a black woman may be
114 Gender as a social category
more likely to be perceived as black if she is in a group dominated by
white people, but more likely to be seen as a woman in a group
dominated by men.
(p. 85)

Extensive research evidence supports the context-dependent nature of

intergroup processes, especially among adults. Contexts that make group
membership diagnostic (e.g., through the actual or anticipated presence of
an out-group) or accessible (e.g., by priming group membership through
labelling) lead to an intensification of intergroup processes. These context
effects have been shown to influence perceptions of similarity between group
members (Doise, Deschamps, & Meyer, 1978; Wilder, 1984), stereotypical
descriptions of self (Charters & Newcomb, 1958) and others (Doosje,
Haslam, Spears, Oakes, & Koomen, 1998; Hopkins, Regan, & Abell, 1997),
and in-group favouritism (Doise & Sinclair, 1973; Dustin & Davis, 1970;
Janssens & Nuttin, 1976; Rabbie & Wilkens, 1971; Stephenson, Skinner, &
Brotherton, 1976; Turner, 1975).
Relatively few studies have examined the effect of context on intergroup
processes in children. Katz and Seavey (1973) asked elementary-school-aged
students to judge the similarity of pairs of faces that varied in terms of colour
(purple, green) and facial expression (smiling, frowning). When the colour or
expression grouping was made salient through the use of novel labels, per-
ceived differences between the groups increased, at least for white children.
Bigler et al. (1997) found that children divided into “blue” and “yellow”
groups designated by coloured tee shirts showed more in-group favouritism in
their attribution of positive and negative traits if their teacher made the
group distinction salient (e.g., by having the children sit according to group
membership or by using verbal group labels).
It seems likely that the salience of gender would vary considerably with
context as well. Deaux and Major (1987) suggest that some events (e.g., having
just watched the Miss America pageant; participating in a mixed-sex group)
may prime us to focus on the gender of other people. Indeed, numerous studies
have shown that the extent to which adults show own-sex favouritism (Hogg
& Turner, 1987; McKillip, DiMiceli, & Luebke, 1977; Schmitt, Silvia, &
Branscombe, 2000; Starer & Denmark, 1974; Todor, 1980) or use gender
stereotypes when describing themselves (Hogg & Turner, 1987; Onorato &
Turner, 1996, 1997, as cited in Turner & Onorato, 1999) or others (McKillip
et al., 1977) increases in contexts designed to enhance the salience of gender.
A few studies have examined the impact of contextual variations on
gender-based intergroup processes among children. Thorne (1987) describes
an incident in which a group of fourth-graders, who were typically quite
gender-segregated on the playground, came together in a mixed-sex group to
defend and discuss the plight of a classmate who had been unjustly punished
by a teacher. The children united on the basis of classroom membership,
which temporarily appeared to become more salient than gender.
 Powlishta 115
Quantitative research evidence supports the importance of context as well.
In one of my own studies (Powlishta, 2002a), boy–girl pairs depicted in
photographs were seen as more similar to each other when they were sur-
rounded by photos of adults and explicitly labelled as children than when the
very same pairs were surrounded by photographs of other children. Focusing
attention on a characteristic that boys and girls share (i.e., childhood)
appeared to make gender less salient, reducing children’s tendency to see
males and females as different sorts of creatures. Sani and Bennett (2001)
similarly found that children attributed different traits to their gender-based
in-group (“boys” or “girls”) depending on whether they had first been asked
to describe an age-based out-group (i.e., a same-sex adult) or a gender-based
out-group (i.e., a child of the other sex). Bigler (1995) also showed that gen-
der stereotyping can vary with context. A subgroup of children whose
teachers made gender salient (e.g., through seating arrangements and label-
ling) were more likely than those in a control condition to view occupations in
a stereotypical fashion.
Finally, at least one study has examined the impact of context on children’s
own-sex favouritism (Deschamps & Doise, 1978). Groups of six boys and six
girls were seated so that each sex occupied two adjacent sides of a table. For
half of the groups, three of the boys and three of the girls were given blue
pens and labelled the “blue group” with the remaining children given red pens
and labelled the “red group”. Thus, for this crossed-category condition there
were two potential ways of making an in-group–out-group distinction (boys
vs. girls; blues vs. reds). The other half of the twelve-person groups were not
further divided (the simple category condition), so that gender was the only
salient basis for group formation. Each child then individually completed a
series of paper-and-pencil games and estimated how well the other children
did at these games. Only in the simple category condition did children predict
that individuals of their own sex had done better than the other sex. In
other words, a context designed to reduce the salience of gender through the
introduction of an alternative basis for categorisation (blue vs. red) reduced
gender-based in-group favouritism.

INDIVIDUAL AND GROUP DIFFERENCES IN GENDER-

BASED INTERGROUP PROCESSES

Gender schematicity
In addition to varying with context, the salience of gender may also vary
from person to person. Although not focusing on individual differences,
self-categorisation theory (Turner & Onorato, 1999) nevertheless proposes
that one factor determining the salience of a particular social category is its
accessibility, that is, the readiness of a person to use a particular categorisa-
tion. In addition to varying with the situation, such accessibility or readiness
is thought to depend on past experience, which can influence the extent
116 Gender as a social category
to which a person identifies with a group or sees the group as central or
valued.
With respect to gender, Bem (1981) similarly proposed that, because
of past exposure to differing social environments, gender should be more
chronically salient for some individuals than for others. In other words,
people vary in their level of gender schematicity, that is, their readiness to
encode and organise information in a gender-based manner (Martin &
Halverson, 1981).
Research evidence has revealed individual differences in gender schematic-
ity as early as the preschool years. Levy and colleagues (Carter & Levy, 1988;
Levy, 1994; Levy & Carter, 1989) measured gender schematicity by asking
children to indicate their toy preferences when presented with pairs of toys.
Those who displayed longer reaction times when asked to choose between
two masculine or two feminine toys, or who displayed shorter reaction times
when asked to choose between a masculine and a feminine toy, relative to
other choice options, were considered to be highly gender schematic. Com-
pared to low-schematic children, those with high-gender schematicity had
poorer memory for stereotype-inconsistent information, were more likely
to make memory errors that distorted information so as to be stereotype-
consistent, showed more gender-typed toy preferences, were more likely to
attribute activities to males and females in a stereotypic fashion, and showed
a greater improvement in recall when items to be remembered shifted from
same-sex gender-typed toys to gender-neutral animals (i.e., “release from
proactive interference”). In other words, not only are some individual chil-
dren more gender-schematic than others, but such schematicity or salience
is related to other aspects of gender-role development.
If children’s gender-based attitudes and beliefs are influenced by the acti-
vation of generic intergroup processes, and if there are individual differences
in the salience of gender (and hence the likelihood that or extent to which
these processes are activated) then we might expect to see correlations among
the various component processes. Indeed, such correlations have been found.
In one study, 8- to 10-year-olds were asked to rate unfamiliar boys and girls
depicted in videotaped vignettes in terms of positive, negative, masculine, and
feminine traits, predicted liking, and similarity to self. Results revealed that
individual differences in the tendency to favour members of one’s own sex, to
use gender stereotypes, to emphasise similarities between the self and same-
sex others, and to homogenise the other sex in these ratings were significantly
correlated with each other (Powlishta, 1995c).
Among adults, there is even more direct evidence that individual differ-
ences in gender salience or schematicity are related to gender-based inter-
group processes. In an intergroup context, female college students with high
levels of gender group identification gave more favourable evaluations
of a same-sex target than did students with low levels of gender group
identification (Schmitt et al., 2000).
 Powlishta 117
Self-esteem
Another individual difference variable related to intergroup processes, and in
particular in-group favouritism, is self-esteem. According to social identity
theory (Tajfel & Turner, 1979), in-group favouritism arises from a desire to
achieve a positive social identity in order to enhance or maintain self-esteem.
Although this theory has been used to predict both negative correlations (low
self-esteem motivates in-group bias) and positive correlations (successful
intergroup discrimination enhances self-esteem) between self-esteem and in-
group favouritism, a recent meta-analysis revealed more support for the latter
prediction. That is, among adults, high self-esteem individuals tend to exhibit
more pro-in-group bias than do low self-esteem individuals (Aberson, Healy,
& Romero, 2000).
This pattern is seen among children as well. Bigler and colleagues found
that when classmates were assigned to “blue” or “yellow” groups, those with
high self-esteem were more likely than those with low self-esteem to favour
their own groups in their positive and negative trait attributions (Bigler et al.,
1997).
If children’s own-sex favouritism results, at least in part, from the generic
tendency to positively evaluate in-groups, then such favouritism should be
related to self-esteem in a similar manner. This hypothesis was tested in a
recent study involving both children and early adolescents (Powlishta &
Vartanian, 1999). Own-sex favouritism was assessed in two ways: the extent
to which participants gave higher liking ratings to same-sex than to other-sex
classmates and the extent to which they attributed positive traits to their own
sex more than to the other. Self-esteem was assessed using Harter’s Self-
Perception Profile for Children (Harter, 1982), which measures perceptions of
one’s own competency in six different domains (scholastic, social acceptance,
athletic, physical appearance, behavioural conduct, and global self-worth).
Self-esteem (in the domain of social acceptance) was positively related to
own-sex favouritism on both measures as predicted, but only for girls.

Sex differences
The previous finding highlights another common and somewhat puzzling
pattern in the literature on children’s gender-based intergroup processes. Not
only are there individual differences in the extent to which these processes are
seen, but there are also fairly consistent sex differences. In general, both sexes
display own-sex favouritism, but girls display stronger biases than do boys.
These sex differences are seen using a wide variety of assessment techniques.
For example, in the study just described, girls favoured their own sex more
than did boys on both measures (liking of classmates, attribution of evalu-
ative traits), and the two favouritism measures were significantly correlated
with each other only for girls (Powlishta & Vartanian, 1999). Other studies
also have found greater own-sex favouritism among girls than boys in their
118 Gender as a social category
attribution of traits to the sexes in general (Egan & Perry, 2001; Parish &
Bryant, 1978; Powlishta, 1995a; Powlishta et al., 1994; Silvern, 1977) or to
individual unfamiliar boys and girls depicted in photographs or videotapes
(Heyman, 2001; Powlishta, 1995c; Zalk & Katz, 1978). Yee and Brown (1994)
found greater bias among girls than boys when asking children to rate how
they felt about each sex, to name “nice” and “mean” things about each sex,
and to award prizes to groups of girls and boys who allegedly had made
collages varying in quality. Girls also display more own-sex favouritism when
naming classmates they dislike or when rating their predicted liking of
unfamiliar peers (Powlishta et al., 1994). Only when it comes to time spent
with same-sex playmates (i.e., gender segregation) do boys often show
stronger own-sex preferences than girls (Maccoby, 1998).
Although most sex differences in gender-based intergroup processes
involve own-sex favouritism, there is some indication that such processes are
stronger more generally for girls than for boys. In a meta-analysis, girls were
found to be more knowledgeable about gender stereotypes than were boys
(Signorella, Bigler, & Liben, 1993). Preliminary evidence indicates that girls
also may show a greater tendency to homogenise the other sex (Powlishta,
1995c) and to encode the gender connotation of words in memory (Perez &
Kee, 2000).
One of the most likely explanations for sex differences in gender-based
intergroup processes, particularly own-sex favouritism, has to do with status
or power differences between males and females (Glick & Hilt, 2000;
Powlishta et al., 1994). Children know that males are considered to be of
higher status than females (Glick & Hilt, 2000; Lockheed & Klein, 1985; see
also David, Grace, & Ryan, Chapter 5). In fact, the first trait stereotypes they
learn in the preschool years portray males as powerful and females as fearful
and helpless (Ruble & Martin, 1998). Perhaps reflecting this difference, mas-
culine stereotyped traits are considered more adult-like/less child-like than
are feminine stereotyped traits by both children and adults (Powlishta, 2000).
In addition, boys are often resistant to influence attempts by girls, enabling
them to dominate cross-sex interactions (Charlesworth & LaFreniere, 1983;
Powlishta & Maccoby, 1990; Serbin, Sprafkin, Elman, & Doyle, 1984). This
dominance may lead to resentment of the other sex among girls (Glick &
Hilt, 2000). More generally, lower status groups show heightened in-group
favouritism under some circumstances, particularly on dimensions not dir-
ectly related to the status difference (Brown, 2000; Van Knippenberg, 1984).
The elevated own-sex favouritism in girls, then, may be a direct reaction to
their lower status and power.
A second possible cause of the sex difference in own-sex favouritism is that
boys may be more concerned with gender stereotypicality than with own-sex
favouritism. Indeed, children are well aware that it is considered worse to be a
sissy than a tomboy (Levy et al., 1995; Martin, 1990), perhaps explaining in
part why gender-typed interests and occupational aspirations tend to become
more flexible for girls but not boys in later childhood (Powlishta et al., 2001).
 Powlishta 119
Adults, too, are particularly unwilling to attribute feminine stereotyped traits
to males (Powlishta, 2000). Hence, boys may be willing to accept traditionally
masculine characteristics even if they are negative and to reject traditionally
feminine characteristics even if they are positive, yielding less apparent own-
sex favouritism. Once again, status differences between the sexes may account
for the differential willingness of boys and girls to deviate from gender norms.
A third factor that may account for the sex differences in own-sex favourit-
ism is acceptance of the overarching stereotypes that girls are “sugar and
spice and everything nice” whereas boys are “snakes and snails and puppy
dog tails” (Serbin et al., 1993). Consistent with this interpretation, Serbin et
al. found that children had greatest knowledge of positive feminine (e.g.,
gentle) and negative masculine (e.g., fights) trait stereotypes, intermediate
knowledge of negative feminine stereotypes (e.g., weak), and least knowledge
of positive masculine stereotypes (e.g., adventurous). In fact, it is even pos-
sible that boys may be willing to attribute negative characteristics to their own
sex because they believe it is “cool to be bad”. Heyman (2001) also found
evidence for a “boys are bad” bias. When shown photographs of unfamiliar
peers described as having performed ambiguous behaviours, children were
more likely to interpret and remember the behaviours in an unfavourable way
when the peer was male rather than female. Both boys and girls showed this
pattern, even though boys predicted they would like the male peers more than
did girls.
A fourth factor that may contribute to the greater own-sex favouritism
typically seen in girls than in boys is a sex difference in the salience of gender
as a grouping characteristic. Yee and Brown (1994) suggested that such a
difference could account for the greater gender bias shown by girls in
selecting prizes for group products, such that boys were rewarding prizes
based on performance (i.e., the best prize to the best product) whereas girls
were rewarding prizes based on gender (i.e., the best prize to girls, regardless
of performance). In other words, boys may have been more attuned to a
performance schema and girls to a gender schema.
Evidence for sex differences in schema activation can be seen in studies
using a release from proactive interference paradigm. As noted earlier, Perez
and Kee (2000) found that girls but not boys seemed to encode the gender
connotation of words in memory. This finding was consistent with an earlier
study of adults (Mills & Tyrrell, 1983) in which women consistently displayed
evidence of encoding the gender connotation of occupations. Men, however,
showed release from proactive interference when shifting from feminine to
masculine occupations but not vice versa, suggesting that masculine occupa-
tions were treated simply as “occupations” unless a gender schema was first
activated by the presentation of feminine items.
Using a different paradigm, Hurtig and Pichevin (1990) also found evi-
dence for asymmetry in gender salience. Adults were asked to describe
another person using as few cues as possible. Gender was more often used as
a cue when the target was a woman rather than a man and when the target
120 Gender as a social category
was seen in a traditionally feminine rather than a neutral situation. Further-
more, female participants more often used gender as the first cue than did
male participants. The authors proposed that the greater salience of “female”
than of “male” results from status differences between the sexes. Because the
higher-status “male” is the default or generic group, it is less likely to cause
gender-schema activation (also see David et al., Chapter 5). Once again, if
gender is more salient for females than for males, this pattern may explain the
greater own-sex favouritism seen among girls.
One final potential explanation for girls’ greater own-sex favouritism has to
do with the way in which favouritism has typically been measured. Perhaps
trait attributions and liking ratings do a better job of capturing how in-group
biases are displayed for girls than for boys. In one minimal group study of
adults, women favoured the in-group by giving more favourable interpersonal
evaluations to in-group members whereas men favoured the in-group by
overevaluating in-group products (Dion, 1979). Consistent with this pattern,
when Deschampes and Doise (1978) used performance predictions as an
index of own-sex favouritism (i.e., asking children to guess how well their
male and female classmates had performed on a novel task), both boys and
girls showed the expected gender biases. However, in this study (unlike most
others) boys were more biased than girls. Additional research including both
trait/liking ratings and performance evaluations to measure children’s
own-sex favouritism is warranted.
Nevertheless, based on the research to date, it appears that girls are more
generally prone to gender-based intergroup processes than are boys. If so,
then why do boys show just as much if not more gender segregation in their
choice of playmates? One possibility is that they are reacting to behavioural
differences between males and females more than to the sex of the playmate,
per se. That is, boys may be seeking out playmates who have similar or com-
patible play styles to their own (e.g., a similar enjoyment of roughhousing);
because of sex differences in play style, such playmates may frequently end up
being other boys (Maccoby, 1998; Powlishta, 1995b; Serbin et al., 1994). Boys
also may avoid playing with other girls for fear of being seen as feminine.

Age/Developmental differences
In addition to individual and sex differences, there may also be age or devel-
opmental differences in the activation of gender-based intergroup processes.
First, there is reason to suspect that gender is particularly salient during early
childhood. As noted previously, except in infancy, gender is almost always
visually apparent; it represents a stable, dichotomous, and exhaustive way of
classifying the self and others. Because young children tend to focus on con-
crete, external attributes and have difficulty dealing with multiple classifica-
tions simultaneously, they may attend to the simplest, most salient basis for
classification, such as gender, rather than to more complex classification
systems. They also may be particularly attentive to gender when they are first
 Powlishta 121
trying to learn about gender roles. As children become older, they are more
likely to focus on internal attributes. Mastering multiple classification skills
should enable them to make finer differentiations among people and to real-
ise that people can simultaneously be “the same” in some ways (e.g., gender)
and “different” in others (e.g., interests, abilities, traits). Attention to gender
also may become less important as gender role expertise is developed. As a
result, gender may become a less salient basis for classification in later child-
hood (Martin & Halverson, 1981; Powlishta et al., 1994; Serbin & Sprafkin,
1986).
Indeed, research supports the notion that gender salience declines some-
what with age (Serbin & Sprafkin, 1986). For example, Yee and Brown (1994)
found that 5-year-olds were more likely to sort photographs on the basis of
gender than were either younger or older children. In another study described
previously, Bennett et al. (2000) attributed a series of statements to boys and
girls depicted in photographs, and then asked children to recall who said
what. In general, children made more within-gender errors (i.e., attributed a
statement made by one boy to another boy or a statement made by one girl to
another girl) than between-gender errors, indicating that they had categorised
the information according to gender. This pattern was stronger for younger
than for older children, perhaps suggesting that gender was more salient for
the younger group. In a study comparing children and adults, when asked to
rate the similarity of pairs of photographed people varying in terms of gen-
der, age, and facial expression, both age groups attended to all three dimen-
sions. However, the salience of gender and age was lower and the salience of
facial expression was higher for adults than for children (Powlishta, 2002b).
In addition to this decline in gender salience, there is a second reason that
there may be developmental changes in the activation of gender-based inter-
group processes. Young children have been shown to display more favouritism
towards experimentally created minimal groups than have older children,
with such favouritism peaking around 5 years of age (Yee & Brown, 1992). To
the extent that generic intergroup processes influence children’s attitudes
towards males and females, then young children may be particularly prone to
own-sex favouritism as well.
Consistent with this decline in gender salience and in the general tendency
to show in-group favouritism, as noted above, a number of studies have found
a reduction in gender-based intergroup processes in later childhood. For
example, although knowledge of gender stereotypes increases and reaches
ceiling, older children (especially girls) often become more flexible in their use
of stereotypes. By the end of the elementary-school-aged years, children
more often acknowledge that “both” males and females can engage in similar
activities and possess similar traits. The inferences they make about other
people become more flexible as well, increasingly influenced by individuating
information (e.g., current behaviour, stated preferences or traits, labels such
as “tomboy” or “sissy”). That is, older children predict that a person with one
feminine characteristic will have others as well, even if the person is a boy,
122 Gender as a social category
whereas a person with one masculine characteristic is predicted to have
others, even if the person is a girl (see Powlishta et al., 2001).
Evidence suggests that this age-related decline in gender stereotyping is tied
to cognitive development. Bigler (1995) found that even after controlling for
age, children who were less able to classify people along multiple dimensions
(e.g., gender, age, race) were more stereotypical in their beliefs about whether
men, women, or both should perform various occupations. And when placed
in an experimental condition in which teachers emphasised gender grouping
in the classroom, only children with poor classification skills gave more
extreme gender stereotypical ratings of their classmates relative to a control
condition. Of course, this doesn’t mean that the use of gender stereotypes
disappears with age; even adults make stereotypical inferences, especially in
their perceptions of children (Powlishta, 2000).
Age patterns for own-sex favouritism are somewhat less consistent than
they are for stereotyping. Although preference for same-sex playmates
remains strong or even increases from early to late childhood (Maccoby,
1998; Powlishta, 2001; Powlishta et al., 1994, 2001; Serbin et al., 1993), other
measures of own-sex favouritism, such as the attribution of positive traits to
one’s own sex, typically decline with age as expected during this period
(Parish, Bryant, & Prawat, 1977; Powlishta et al., 1994; Zalk & Katz, 1978).
Not surprisingly, the decline in own-sex favouritism continues into ado-
lescence. Several studies have found a reduced tendency to attribute more
favourable traits to one’s own sex in adolescence relative to childhood (Egan
& Perry, 2001; Parish & Bryant, 1978; Powlishta & Vartanian, 1999). In fact,
in one study, adolescent boys actually rated the other sex more favourably
than their own (Parish & Bryant, 1978).Young adolescents show less own-sex
favouritism than children in their liking of classmates as well; but despite
these reductions, even early adolescents continue to show some degree of
own-sex favouritism, both in their general perception of the sexes and in their
attitudes towards familiar individuals (Powlishta & Vartanian, 1999). Glick
and Hilt (2000) suggest that in adolescence, the hostile intergroup attitudes
characteristic of childhood do not simply disappear, but are combined with a
new “benevolent sexism”, particularly on the part of boys towards girls. This
benevolent sexism is a patronising form of prejudice in which perceivers
believe that their attitudes towards the target group are favourable, but at the
same time view the other group as inferior to their own.

HOW IS GENDER DIFFERENT FROM OTHER

SOCIAL CATEGORIES?

The decline in own-sex favouritism seen during adolescence highlights one

important difference between gender and other social categories. As Williams
and Giles (1978) point out, the “psychological and physiological dependen-
cies between men and women, their numerical balance in the population and
 Powlishta 123
their positive affect for each other” (p. 431) distinguish gender from many
other social groupings. Given that taboos against cross-sex interaction
start to weaken in most Western societies between 11 and 13 years of age
(Maccoby, 1998), it is not surprising that gender-based biases decline in early
adolescence as sexual attraction begins to compete with own-sex favouritism
(at least for heterosexual individuals).
Own-sex favouritism does not disappear even in adulthood, however. It
does become more unpredictable, though, in that in some studies people
favour their own sex (Starer & Denmark, 1974) and in others they show a
general pro-male or pro-female bias or no bias at all (McKillip et al., 1977;
Todor, 1980). Such variations exemplify the ambivalence and complexity that
characterises adolescent and adult gender relations (Glick & Hilt, 2000). The
discrepancies in research findings may be accounted for, at least in part, by
contextual variations that heighten the salience of gender in-groups and out-
groups (leading to own-sex favouritism or traditional stereotyping) versus
sexuality (leading to more benevolent attitudes towards the other sex). Con-
sistent with this interpretation, Hogg and Turner (1987) found that female
participants favoured men in a point allocation task when tested in mixed-sex
dyads (which might be expected to make sexuality salient) but favoured
women when tested in mixed-sex groups (which might be expected to make
gender in-groups and out-groups salient).
The discrepancy in findings of own-sex favouritism among adults may also
be accounted for by the influence of gender stereotypes. For example, when
women were asked to evaluate products that might be considered tradition-
ally masculine (“published journal articles” on various academic topics), they
gave more favourable ratings when the products were attributed to male
rather than female authors (Goldberg, 1968). But when women read articles
on more traditionally feminine topics (marriage, child discipline, and special
education), there was a nonsignificant trend to evaluate the “female-
authored” articles more highly than the very same “male-authored” articles
(see Pheterson, Kiesler, & Goldberg, 1971).
These latter findings point to a second factor that sets gender apart from
other social categories, and particularly from the sorts of minimal groups
created in laboratory settings. Existing stereotypes (positive and negative)
about each sex and actual behavioural differences between the sexes can
influence children’s attitudes and behaviours in ways that sometimes run
counter to and sometimes enhance generic intergroup processes. We have
already seen that children have extensive knowledge about traditional gender
stereotypes. Some of these stereotypes portray their own sex in a negative
fashion. For example, boys are aware that males are considered to be messier
than females; girls are aware that females are considered to be weaker than
males. Acknowledging these stereotypes reduces the appearance of own-sex
favouritism. Yet if examined carefully, the influence of in-group bias can
nevertheless be seen in children’s gender stereotyped attitudes. Girls believe
that traits like messiness are very masculine whereas boys see them as only
124 Gender as a social category
somewhat so; likewise, boys view traits such as weakness as more feminine
than do girls. The reverse pattern is seen for positive gender-stereotyped
traits, where each sex will acknowledge that the other has positive stereo-
typical characteristics (e.g., that girls are gentle and boys are strong), yet have
more extreme beliefs about the positive own-sex stereotypes (Powlishta,
1995a). In other words, just as with national groups (Bennett, Lyons, Sani, &
Barrett, 1998), children’s gender-related attitudes may reflect the expression
of socially shared knowledge (i.e., stereotypes) in addition to any generic
in-group bias.
In fact, as with national groups where children sometimes show in-group
favouritism even before subjective identification with the in-group has
occurred (Bennett et al., 1998), very young children may display gender-
segregated play before they are fully aware of the relationship between sex of
self and others (Maccoby, 1998). This pattern could arise from biologically
and/or environmentally determined sex differences in play styles, coupled
with a preference for playmates with similar or compatible styles. In other
words, as noted earlier, children may not always be reacting to the sex of a
potential playmate per se, but instead to the sorts of positive experiences they
have had during face-to-face contacts with individual children of their own
sex. That is, children’s same-sex preferences are probably influenced by bio-
logical and social-environmental factors as well as by generic intergroup
processes.
Perceived differences between and similarities within the sexes are also
most likely determined by more than generic intergroup processes. Children
encounter models of gender-typed behaviour in their homes, schools, and the
media (see Powlishta et al., 2001). For example, despite changes in recent
years, men and women often continue to divide household and child-care
tasks along traditional gender lines. Men are more likely to work, and work
longer hours, outside of the home. Women are more often teachers and men
more often administrators in elementary schools. Stereotypical portrayals are
perhaps even more extreme in television programmes, commercials, video
games, and books. When children describe males and females as being differ-
ent, then, they may not simply be exaggerating intragroup similarities and
intergroup differences. Instead, as Glick and Hilt (2000) put it, “Gender-
related attitudes exaggerate, but also reflect (and help to reinforce), social
realities” (p. 266).
In sum, gender groups are different from experimentally created minimal
groups in that males and females do have pre-existing differences, experiences,
hostilities (or attractions), and face-to-face contacts. Thus, although
generic intergroup processes most likely contribute to children’s beliefs and
preferences about the sexes, such processes do not fully explain gender-role
development. Cognitive-developmental, social-environmental, and biological
factors must also be taken into account. The complex patterns seen in child-
ren’s gender-related attitudes and behaviours also have broader implications
for those studying intergroup processes. In particular, the social identity
 Powlishta 125
approach could give greater attention to the fact that social categories differ
from one another in ways that may have significant consequences, especially
in a developmental context.

CONCLUSIONS AND IMPLICATIONS FOR IMPROVING

CROSS-SEX RELATIONS

In summary, the same phenomena that are seen in social psychological

studies of natural and experimentally created groups of adults are found with
respect to children’s attitudes and beliefs about males and females. Children
favour their own sex in their choice of playmates, in their liking and trait
ratings of males and females both individually and globally, and in their
evaluations of male and female performance. In other words, consistent with
social identity theory, children believe that their own sex is better than the
other. At least for girls, such own-sex favouritism is linked to heightened
self-esteem, as expected.
This in-group favouritism is accompanied by other intergroup phenomena.
Namely, children exaggerate differences between and similarities within the
gender groups, show a readiness to learn and create new gender stereotypes,
and resist stereotypic changes or violations. Although not yet studied exten-
sively, there also may be a tendency to view the other sex as particularly
homogeneous. These distorted perceptions of variability are correlated with
own-sex favouritism, consistent with a common origin in gender-based social
categorisation.
As predicted by self-categorisation theory, these gender-based intergroup
processes are elevated in contexts designed to enhance the salience of gender
and among individuals for whom gender seems to be chronically salient. Such
processes tend to be stronger in girls than in boys, perhaps reflecting the sort
of status effects that have been found with experimentally created groups as
well. And despite the sexual attraction and increased cognitive maturity that
is most likely to account for the decline in gender-based intergroup processes
with age, even adults sometimes favour their own sex and make use of gender
stereotypes.
These findings suggest that a tendency to categorise the world into
male and female groups, viewing same-sex individuals as “us” as opposed
to “them”, contributes to children’s gender-role development above and
beyond any domain-specific processes (e.g., biological sex differences,
gender-differentiated socialisation). Children may view their own sex as
superior in order to achieve a positive social identity, which in turn should
motivate them to adopt gender-typed characteristics and prefer same-sex
playmates. The resulting gender-segregated play exposes boys and girls
to different socialisation contexts, potentially creating or amplifying sex
differences (Martin & Fabes, 2001). The generic tendency to exaggerate
between-group differences and within-group similarities may contribute to
126 Gender as a social category
gender stereotypes, further encouraging boys and girls to adopt different
behaviours and roles.
This intergroup perspective on gender role development has implications
for how one might encourage boys and girls to have more positive attitudes
about each other. Rather than attempting to change children’s behaviours
and beliefs directly, for example by rewarding cross-sex interactions (Serbin,
Tonick, & Sternglanz, 1977) or by providing training that contradicts gender
stereotypes (Bigler & Liben, 1992), a more promising approach may be to
inhibit the formation of gender-based categories. Of course, one cannot
block the basic human tendency to form social categories that help us make
sense of the world and feel good about ourselves. And we have seen that,
relative to other potential grouping characteristics, gender is a rather chron-
ically salient basis for social categorisation among young children. However,
we also have seen that it is possible to create contexts that focus attention on
characteristics other than gender, for example situations that require the
cooperation of boys and girls to achieve goals or that introduce explicit
groupings that cross-cut gender. By encouraging children to categorise them-
selves, at least temporarily, in some manner other than gender, this approach
shows promise for reducing children’s biases against the other sex and their
tendency to view each other in a rigidly stereotypical fashion.

REFERENCES

Aberson, C. L., Healy, M., & Romero, V. (2000). Ingroup bias and self-esteem: A
meta-analysis. Personality and Social Psychology Review, 4, 157–173.
Albert, A. A., & Porter, J. R. (1983). Age patterns in the development of children’s
gender-role stereotypes. Sex Roles, 9, 59–67.
Allport, G. (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley.
Bauer, P. J., & Coyne, M. J. (1997). When the name says it all: Preschoolers’ recogni-
tion and use of the gendered nature of common proper names. Social Development,
6, 271–291.
Bauer, P. J., Liebl, M., & Stennes, L. (1998). Pretty is to dress as brave is to
suitcoat: Gender-based property-to-property inferences by 4½-year-old children.
Merrill-Palmer Quarterly, 44, 355–377.
Beere, C. A. (1990). Gender roles: A handbook of tests and measures. New York:
Greenwood.
Bem, S. L. (1981). Gender schema theory: A cognitive account of sex typing.
Psychological Review, 88, 354–364.
Bennett, M., Lyons, E., Sani, F., & Barrett, M. (1998). Children’s subjective indentifi-
cation with the group and in-group favouritism. Developmental Psychology, 34,
902–909.
Bennett, M., Sani, F., Hopkins, N., Agostini, L., & Malucchi, L. (2000). Children’s
gender categorization: An investigation of automatic processing. British Journal of
Developmental Psychology, 18, 97–102.
Berndt, T. J., & Heller, K. A. (1986). Gender stereotypes and social inferences: A
developmental study. Journal of Personality and Social Psychology, 50, 889–898.
 Powlishta 127
Best, D. L. (1982). An overview of findings from children studies of sex-trait stereo-
types in 23 countries. In R. Rath, H. S. Asthana, D. Sinha, & J. B. H. Sinha (Eds.),
Diversity and unity in cross-cultural psychology (pp. 261–271). Amsterdam: Swets.
Best, D. L., Williams, J. E., Cloud, J. M., Davis, S. W., Roberston, L. S., Edwards, J. R.,
Giles, H., & Fowles, J. (1977). Development of sex-trait stereotypes among
young children in the United States, England, and Ireland. Child Development, 48,
1375–1384.
Biernat, M. (1991). Gender stereotypes and the relationship between masculinity and
femininity: A developmental analysis. Journal of Personality and Social Psychology,
61, 351–365.
Bigler, R. S. (1995). The role of classification skill in moderating environmental
influences on children’s gender stereotyping: A study of the functional use of gen-
der in the classroom. Child Development, 66, 1072–1087.
Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorization and the
formation of intergroup attitudes in children. Child Development, 68, 530–543.
Bigler, R. S., & Liben, L. S. (1992). Cognitive mechanisms in children’s gender stereo-
typing: Theoretical and educational implications of a cognitive-based intervention.
Child Development, 63, 1351–1363.
Blaske, D. M. (1984). Occupational sex-typing by kindergarten and fourth-grade
children. Psychological Reports, 54, 795–801.
Brewer, M. B., Ho, H., Lee, J., & Miller, N. (1987). Social identity and social distance
among Hong Kong school children. Personality and Social Psychology Bulletin, 13,
156–165.
Brown, R. (2000). Social identity theory: Past achievements, current problems and
future challenges. European Journal of Social Psychology, 30, 745–778.
Cann, A., & Palmer, S. (1986). Children’s assumptions about the generalizability of
sex-typed abilities. Sex Roles, 15, 551–558.
Carter, D. B., & Levy, G. D. (1988). Cognitive aspects of early sex-role development:
The influence of gender schemas on preschoolers’ memories and preferences for
sex-typed toys and activities. Child Development, 59, 782–792.
Carter, D. B., & Patterson, C. J. (1982). Sex roles as social conventions: The develop-
ment of children’s conceptions of sex-role stereotypes. Developmental Psychology,
18, 812–824.
Charlesworth, W. R., & LaFreniere, P. (1983). Dominance, friendship, and resource
utilization in preschool children’s groups. Ethology and Sociobiology, 4, 603–615.
Charters, W. W., & Newcomb, T. M. (1958). Some attitudinal effects of experimentally
increased salience of a membership group. In E. E. Maccoby, T. M. Tewcomb, & E.
L. Hartley (Eds.), Readings in social psychology (pp. 276–281). New York: Henry
Holt.
Cordua, G. D., McGraw, K. O., & Drabman, R. S. (1979). Doctor or nurse: Children’s
perception of sex typed occupations. Child Development, 50, 590–593.
Cowan, G., & Hoffman, C. D. (1986). Gender stereotyping in young children:
Evidence to support a concept-learning approach. Sex Roles, 14, 211–224.
Damon, W. (1977). The social world of the child. San Francisco: Jossey-Bass.
Deaux, K., & Major, B. (1987). Putting gender into context: An interactive model of
gender-related behavior. Psychological Review, 94, 369–389.
Deschamps, J. C., & Doise, W. (1978). Crossed category memberships in intergroup
relations. In H. Tajfel (Ed.), Differentiation between social groups (pp. 141–158).
London: Academic Press.
128 Gender as a social category
Dion, K. L. (1979). Status equity, sex composition of group, and intergroup bias.
Personality and Social Psychology Bulletin, 5, 240–244.
Doise, W., Deschamps, J.-C., & Meyer, G. (1978). The accentuation of intra-category
similarities. In H. Tajfel (Ed.), Differentiation between social groups (pp. 159–168).
London: Academic Press.
Doise, W., & Sinclair, A. (1973). The categorisation process in intergroup relations.
European Journal of Social Psychology, 3, 145–157.
Doosje, B., Haslam, S. A., Spears, R., Oakes, P. J., & Koomen, W. (1998). The effect
of comparative context on central tendency and variability judgments and the
evaluation of group characteristics. European Journal of Social Psychology, 28,
173–184.
Dustin, D. S., & Davis, H. P. (1970). Evaluative bias in group and individual
competition. European Journal of Social Psychology, 80, 103–108.
Edelbrock, C., & Sugawara, A. I. (1978). Acquisition of sex-typed preferences in
preschool-aged children. Developmental Psychology, 14, 614–623.
Egan, S. K., & Perry, D. G. (2001). Gender identity: A multidimensional analysis with
implications for psychosocial adjustment. Developmental Psychology, 37, 451–463.
Eichstedt, J. A., Serbin, L. A., Poulin-Dubois, D., & Sen, M. G. (in press). Of bears
and men: Infants’ knowledge of conventional and metaphorical gender stereotypes.
Infant Behavior and Development.
Etaugh, C., & Riley, S. (1979). Knowledge of sex stereotypes in preschool children.
Psychological Reports, 44, 1279–1282.
Fagan, J. F., & Shepherd, P. A. (1981). Theoretical issues in the early development
of visual perception. In M. Lewis & L. Taft (Eds.), Developmental disabilities in
preschool children (pp. 9–34). New York: Spectrum.
Fagan, J. F., & Singer, L. T. (1979). The role of simple feature differences in infants’
recognition of faces. Infant Behavior and Development, 2, 39–45.
Gelman, S., Collman, P., & Maccoby, E. (1986). Inferring properties from categories
versus inferring categories from properties: The case of gender. Child Development,
57, 396–404.
Glick, P., & Hilt, L. (2000). Combative children to ambivalent adults: The develop-
ment of gender prejudice. In T. Eckes & H. M. Trautner (Eds.), The developmental
social psychology of gender (pp. 243–272). Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Goldberg, P. (1968). Are women prejudiced against women? Transaction, 5, 28–30.
Harter, S. (1982). The Perceived Competence Scale for Children. Child Development,
53, 87–97.
Haugh, S. S., Hoffman, C. D., & Cowan, G. (1980). The eye of the very young
beholder: Sex typing of infants by young children. Child Development, 51, 598–600.
Hayden-Thomson, L., Rubin, K. H., & Hymel, S. (1987). Sex preferences in
sociometric choices. Developmental Psychology, 23, 558–562.
Helwig, A. A. (1998). Gender-role stereotyping: Testing theory with a longitudinal
sample. Sex Roles, 38, 403–423.
Heyman, G. D. (2001). Children’s interpretation of ambiguous behavior: Evidence for
a “boys are bad” bias. Social Development, 10, 230–247.
Hogg, M. A., & Turner, J. C. (1987). Intergroup behaviour, self-stereotyping and the
salience of social categories. British Journal of Social Psychology, 26, 325–340.
Hopkins, N., Regan, M., & Abell, J. (1997). On the context dependence of national
stereotypes: Some Scottish data. British Journal of Social Psychology, 36, 553–563.
 Powlishta 129
Hurtig, M.-C., & Pichevin, M.-F. (1990). Salience of the sex category system in person
perception: Contextual variations. Sex Roles, 22, 369–395.
Janssens, L., & Nuttin, J. R. (1976). Frequency perception of individual and
group successes as a function of competition, coaction, and isolation. Journal of
Personality and Social Psychology, 34, 830–836.
Johnston, K. E., Madole, K. L., Bittinger, K., & Smith, A. (2001). Developmental
changes in infants’ and toddlers’ attention to gender categories. Merrill-Palmer
Quarterly, 47, 563–584.
Kail, R. V. Jr., & Levine, L. E. (1976). Encoding processes and sex-role preferences.
Journal of Experimental Child Psychology, 21, 256–263.
Katz, P. A. (1986). Modification of children’s gender-stereotyped behavior: General
issues and research considerations. Sex Roles, 14, 591–602.
Katz, P. A. (1996). Raising feminists. Psychology of Women Quarterly, 20, 323–340.
Katz, P. A., & Seavey, C. (1973). Labels and children’s perception of faces. Child
Development, 44, 770–775.
Kuhn, D., Nash, S., & Brucken, L. (1978). Sex role concepts of two- and
three-year-olds. Child Development, 49, 445–451.
Leinbach, M. D., & Fagot, B. I. (1986). Acquisition of gender labels: A test for
toddlers. Sex Roles, 15, 655–667.
Leinbach, M. D., & Fagot, B. I. (1993). Categorical habituation to male and female
faces: Gender schematic processing in infancy. Infant Behavior and Development, 16,
317–332.
Leinbach, M. D., Hort, B. E., & Fagot, B. I. (1997). Bears are for boys: Metaphorical
associations in young children’s gender stereotypes. Cognitive Development, 12,
107–130.
Levy, G. D. (1994). High and low gender schematic children’s release from proactive
interference. Sex Roles, 30, 93–108.
Levy, G. D. (1999). Gender-typed and non-gender-typed category awareness in
toddlers. Sex Roles, 41, 851–873.
Levy, G. D., & Carter, D. B. (1989). Gender schema, gender constancy, and
gender-role knowledge: The roles of cognitive factors in preschoolers’ gender-role
stereotype attributions. Developmental Psychology, 25, 444–449.
Levy, G. D., & Haaf, R. A. (1994). Detection of gender-related categories by
10-month-old infants. Infant Behavior and Development, 17, 457–459.
Levy, G. D., Taylor, M. G., & Gelman, S. A. (1995). Traditional and evaluative aspects
of flexibility in gender roles, social conventions, moral rules, and physical laws.
Child Development, 66, 515–531.
Livesley, W. J., & Bromley, D. B. (1973). Person perception in childhood and
adolescence. London: John Wiley & Sons.
Lockheed, M., & Klein, S. (1985). Sex equity in classroom organization and
climate. In S. Klein (Ed.), Handbook for achieving sex equity through education
(pp. 189–217). Baltimore, MD: Johns Hopkins University Press.
Maccoby, E. E. (1988). Gender as a social category. Developmental Psychology, 24,
755–765.
Maccoby, E. E. (1998). The two sexes: Growing up apart, coming together. Cambridge,
MA: Harvard University Press.
Maccoby, E. E., & Jacklin, C. N. (1987). Gender segregation in childhood. In H. Reese
(Ed.), Advances in child development and behavior, Vol. 20 (pp. 239–288). New York:
Academic Press.
130 Gender as a social category
Martin, C. L. (1989). Children’s use of gender-related information in making social
judgments. Developmental Psychology, 25, 80–88.
Martin, C. L. (1990). Attitudes and expectations about children with nontraditional
and traditional gender roles. Sex Roles, 22, 151–165.
Martin, C. L. (2000). Cognitive theories of gender development. In T. Eckes & H. M.
Trautner (Eds.), The developmental social psychology of gender (pp. 91–121).
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Martin, C. L., Eisenbud, L., & Rose, H. (1995). Children’s gender-based reasoning
about toys. Child Development, 66, 1453–1471.
Martin, C. L., & Fabes, R. A. (2001). The stability and consequences of young
children’s same-sex peer interactions. Developmental Psychology, 37, 431–446.
Martin, C. L., & Halverson, C. F. Jr. (1981). A schematic processing model of sex
typing and stereotyping in children. Child Development, 52, 1119–1134.
Martin, C. L., & Halverson, C. F. Jr. (1983). The effects of sex-typing schemas
on young children’s memory. Child Development, 54, 563–574.
Martin, C. L., & Little, J. K. (1990). The relation of gender understanding to
children’s sex-typed preferences and gender stereotypes. Child Development, 61,
1427–1439.
Martin, C. L., Wood, C. H., & Little, J. K. (1990). The development of gender
stereotype components. Child Development, 61, 1891–1904.
McKillip, J., DiMiceli, A. J., & Luebke, J. (1977). Group salience and stereotyping.
Social Behavior and Personality, 5, 81–85.
Messick, D. M., & Mackie, D. M. (1989). Intergroup relations. Annual Review of
Psychology, 40, 45–81.
Miller, C. (1983). Developmental changes in male/female voice classification by
infants. Infant Behavior and Development, 6, 313–330.
Mills, C. J., & Tyrrell, D. J. (1983). Sex-stereotypic encoding and release from
proactive interference. Journal of Personality and Social Psychology, 45, 772–781.
Moghaddam, F. M., & Stringer, P. (1986). “Trivial” and “important” criteria for social
categorization in the minimal group paradigm. Journal of Social Psychology, 126,
345–354.
Nadleman, L. (1974). Sex identity in American children: Memory, knowledge and
preference tests. Developmental Psychology, 10, 413–417.
Nesdale, D., & Flesser, D. (2001). Social identity and the development of children’s
group attitudes. Child Development, 72, 506–517.
Parish, T. S., & Bryant, W. T. (1978). Mapping sex group stereotypes of elementary
and high school students. Sex Roles, 4, 135–140.
Parish, T. S., Bryant, W. T., & Prawat, R. S. (1977). Reversing effects of sexism in
elementary school girls through counterconditioning. Journal of Instructional
Psychology, 4, 11–16.
Park, B., & Rothbart, M. (1982). Perception of out-group homogeneity and levels of
social categorization: Memory for the subordinate attributes of in-group and out-
group members. Journal of Personality and Social Psychology, 42, 1051–1068.
Perez, S. M., & Kee, D. W. (2000). Girls not boys show gender-connotation encoding
from print. Sex Roles, 42, 439–447.
Pheterson, G. I., Kiesler, S. B., & Goldberg, P. A. (1971). Evaluation of the
performance of women as a function of their sex, achievement, and personal
history. Journal of Personality and Social Psychology, 19, 114–118.
Poulin-Dubois, D., Serbin, L. A., Eichstedt, J. A., & Sen, M. G. (in press). Men don’t
 Powlishta 131
put on make-up: Toddlers’ knowledge of the gender stereotyping of household
activities. Social Development.
Powlishta, K. K. (1995a). Gender bias in children’s perception of personality traits.
Sex Roles, 32, 17–28.
Powlishta, K. K. (1995b). Gender segregation among children: Understanding the
“cootie phenomenon”. Young Children, May, 61–69.
Powlishta, K. K. (1995c). Intergroup processes in childhood: Social categorization
and sex role development. Developmental Psychology, 31, 781–788.
Powlishta, K. K. (2000). The effect of target age on the activation of gender
stereotypes. Sex Roles, 42, 271–282.
Powlishta, K. K. (2001, April). Own-sex favouritism and gender-role development.
Poster presented at the Biennial Meeting of the Society for Research in Child
Development, Minneapolis, MN.
Powlishta, K. K. (2002a). Perceived similarity of boys and girls: Contextual variations
in the salience of gender. Manuscript submitted for publication.
Powlishta, K. K. (2002b). The salience of gender in social perception. Manuscript
submitted for publication.
Powlishta, K. K., & Maccoby, E. E. (1990). Resource utilization in mixed-sex dyads:
The influence of adult presence and task type. Sex Roles, 29, 723–737.
Powlishta, K. K., Sen, M. G., Serbin, L. A., Poulin-Dubois, D., & Eichstedt, J. A.
(2001). From infancy through middle childhood: The role of cognitive and social
factors in becoming gendered. In R. K. Unger (Ed.), Handbook of the psychology of
women and gender (pp. 116–132). New York: John Wiley & Sons.
Powlishta, K. K., Serbin, L. A., Doyle, A., & White, D. C. (1994). Gender, ethnic,
and body type biases: The generality of prejudice in childhood. Developmental
Psychology, 30, 526–536.
Powlishta, K. K., & Vartanian, L. R. (1999, April). Self-esteem and own-sex
favouritism in middle childhood and early adolescence. Poster presented at the
Biennial Meeting of the Society for Research in Child Development, Albuquerque,
NM.
Quattrone, G. A. (1986). On the perception of a groups’ variability. In S. Worchel &
W. G. Austin (Eds.), Psychology of intergroup relations (pp. 25–48). Chicago, IL:
Nelson-Hall Publishers.
Rabbie, J. M., & Wilkens, G. (1971). Intergroup competition and its effect on intra-
group and intergroup relations. European Journal of Social Psychology, 2, 215–234.
Reis, H. T., & Wright, S. (1982). Knowledge of sex-role stereotypes in children aged 3
to 5. Sex Roles, 8, 1049–1056.
Ruble, D. N., & Martin, C. L. (1998). Gender development. In N. Eisenberg (Ed.), W.
Damon (Series Editor), Handbook of child psychology. Vol. 3: Social, emotional and
personality development (5th ed., pp. 933–1016). New York: John Wiley & Sons.
Rutland, A. (1999). The development of national prejudice, in-group favouritism and
self-stereotypes in British children. British Journal of Social Psychology, 38, 55–70.
Sani, F., & Bennett, M. (2001). Contextual variability in young children’s gender
ingroup stereotype. Social Development, 10, 221–229.
Schmitt, M. T., Silvia, P. J., & Branscombe, N. R. (2000). The intersection of
self-evaluation maintenance and social identity theories: Intragroup judgment in
intergroup and interpersonal contexts. Personality and Social Psychology Bulletin,
26, 1598–1606.
Sen, M. G., & Bauer, P. J. (2002). Cognitive correlates of two-year-old children’s
132 Gender as a social category
differential memory for own-gender-stereotyped information. Manuscript submitted
for publication.
Serbin, L. A., Moller, L. C., Gulko, J., Powlishta, K. K., & Colburne, K. A. (1994).
The emergence of gender segregation in toddler playgroups. In C. Leaper (Ed.),
Childhood gender segregation: Causes and consequences. New directions for child
development, Vol. 65 (pp. 7–17). San Francisco: Jossey-Bass.
Serbin, L. A., Poulin-Dubois, D., Colburne, K. A., Sen, M. G., & Eichstedt, J. A.
(2001). Gender stereotyping in infancy: Visual preferences for and knowledge of
gender-stereotyped toys in the second year. International Journal of Behavioral
Development, 25, 7–15.
Serbin, L. A., Powlishta, K. K., & Gulko, J. (1993). The development of sex-typing in
middle childhood. Monographs of the Society for Research in Child Development, 58
(Serial No. 232).
Serbin, L. A., & Sprafkin, C. (1986). The salience of gender and the process of
sex-typing in three- to seven-year-old children. Child Development, 57, 1188–1199.
Serbin, L. A., Sprafkin, C., Elman, M., & Doyle, A. B. (1984). The early development
of sex differentiation patterns and social influence. Canadian Journal of Social
Science, 14, 350–368.
Serbin, L. A., Tonick, I. J., & Sternglanz, S. H. (1977). Shaping cooperative cross-sex
play. Child Development, 48, 924–929.
Signorella, M. L. (1987). Gender schemata: Individual differences and context effects.
In L. S. Liben & M. L. Signorella (Eds.), Children’s gender schemata (pp. 23–37).
London: Jossey-Bass.
Signorella, M. L., Bigler, R. S., & Liben, L. S. (1993). Developmental differences in
children’s gender schemata about others: A meta-analytic review. Developmental
Review, 13, 147–183.
Silvern, L. E. (1977). Children’s sex-role preferences: Stronger among girls than boys.
Sex Roles, 3, 159–171.
Stangor, C., & Ruble, D. N. (1987). Development of gender role knowledge
and gender constancy. In L. S. Liben & M. L. Signorella (Eds.), Children’s gender
schemata (pp. 5–22). London: Jossey-Bass.
Starer, R., & Denmark, F. (1974). Discrimination against aspiring women.
International Journal of Group Tensions, 4, 65–70.
Stephenson, G. M., Skinner, M., & Brotherton, C. J. (1976). Group participation
and intergroup relations: An experimental study of negotiation groups. European
Journal of Social Psychology, 6, 51–70.
Tajfel, H. (1970). Experiments in intergroup discrimination. Scientific American,
November, 96–102.
Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categorization and
intergroup behaviour. European Journal of Social Psychology, 1, 149–178.
Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In
W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations
(pp. 33–47). Monterey, CA: Brooks/Cole.
Taylor, S. E. (1981). A categorization approach to stereotyping. In D. L. Hamilton
(Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 83–114).
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Thompson, S. K. (1975). Gender labels and early sex role development. Child
Development, 46, 339–347.
Thorne, B. (1987, February). Children and gender: Constructions of difference. Paper
 Powlishta 133
presented at the Conference on Theoretical Perspectives on Sexual Difference,
Stanford University, Stanford, CA.
Todor, N. L. (1980). The effect of the sexual composition of a group on discrimination
against women and sex-role attitudes. Psychology of Women Quarterly, 5, 292–310.
Trautner, H. M., Helbing, N., Sahm, W. B., & Lohaus, A. (1989, April). Beginning
awareness-rigidity-flexibility: A longitudinal analysis of sex-role stereotyping in 4- to
10-year-old children. Paper presented at the meeting of the Society for Research in
Child Development, Kansas City.
Turner, J. (1975). Social comparison and social identity: Some prospects for
intergroup behaviour. European Journal of Social Psychology, 5, 5–34.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorization theory. Oxford: Basil Blackwell.
Turner, J. C., & Onorato, R. (1999). Social identity, personality, and the self-concept:
A self-categorization perspective. In T. R. Tyler, R. M. Kramer, & O. P. John (Eds.),
The psychology of the social self (pp. 11–46). Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Van Knippenberg, A. F. M. (1984). Intergroup differences in group perceptions. In H.
Tajfel (Ed.), The social dimension: European developments in social psychology, Vol.
2 (pp. 560–578). Cambridge: Cambridge University Press.
Vaughan, G., Tajfel, H., & Williams, J. A. (1981). Bias in reward allocation in an
intergroup and an interpersonal context. Social Psychology Quarterly, 44, 37–42.
Vener, A. M., & Snyder, C. A. (1966). The preschool child’s awareness and
anticipation of adult sex-roles. Sociometry, 29, 159–168.
Weinraub, M., Clemens, L. P., Sockloff, A., Ethridge, T., Gracely, E., & Meyers, B.
(1984). The development of sex role stereotypes in the third year: Relationship to
gender labelling, identity, sex-typed toy preference, and family characteristics. Child
Development, 55, 1493–1503.
Wetherell, M. (1982). Cross-cultural studies of minimal groups: Implications for the
social identity theory of intergroup relations. In H. Tajfel (Ed.), Social identity and
intergroup relations (pp. 207–240). Cambridge: Cambridge University Press.
Wilder, D. A. (1981). Perceiving persons as a group: Categorization and intergroup
behavior. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup
behavior (pp. 213–257). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Wilder, D. A. (1984). Predictions of belief homogeneity and similarity following social
categorization. British Journal of Social Psychology, 23, 323–333.
Williams, J. E., Bennett, S. M., & Best, D. L. (1975). Awareness and expression of sex
stereotypes in young children. Developmental Psychology, 11, 635–642.
Yee, M., & Brown, R. (1992). Self-evaluations and intergroup attitudes in children
aged three to nine. Child Development, 63, 619–629.
Yee, M., & Brown, R. (1994). The development of gender differentiation in young
children. British Journal of Social Psychology, 33, 183–196.
Zalk, S. R., & Katz, P. A. (1978). Gender attitudes in children. Sex Roles, 4, 349–357.
5 The gender wars: A
self-categorisation theory
perspective on the development
of gender identity
Barbara David, Diana Grace, and
Michelle K. Ryan

“Who am I?” might be said to be the pre-eminent question of childhood. After

telling us their name, the answer most children will give is that they are a
child, and a girl or a boy (Brooks & Lewis, 1976; Fiske, 1992; Harris, 1995). If
they live in a social context of high ethnic, religious, or national conflict, the
child might also tell us that they are Serbian or Croatian, Catholic or Protest-
ant, Israeli or Palestinian (Hirschfeld, 1993; Stevenson & Stevenson, 1960;
see also Nesdale, Chapter 8, and Verkuyten, Chapter 7).
Empirical evidence confirms that the dimensions of age, gender, and, in
some cases, race, are those on which children’s first social identities are based.
Of these, some researchers have claimed that “gender is the most funda-
mental” (Banaji & Prentice, 1994, p. 315) and there is certainly evidence that
it is one of the earliest of which children show awareness: Infants are able to
make categorical distinctions between males and females even before they
have the language to articulate the differences (Leinbach & Fagot, 1993;
Walker-Andrews, Bahrick, Raglioni, & Diaz, 1991). By the age of 2 to 3
years, most children not only distinguish between males and females but can
tell us that they, themselves, are a boy or a girl, and will exhibit some distress
if an adult assigns them to the wrong sex (Bussey, 1986; Money & Ehrhardt,
1972).
Correctly applying the label for one’s biological sex, however, only answers
the opening question in part for, without a meaning, the label is a long way
from being an identity. It is obtaining that meaning, the process of gender
socialisation, that will be the major focus of this chapter. We will first outline
what is known about the developmental sequence of gender socialisation,
and address the nature of the knowledge being acquired: What does it mean
to be male or female in Western society? Second, we will outline and evaluate
the major theoretical approaches. In the third section we will explicate what
we see as the fundamental difference between an understanding of the self as
an individual product – the concept underlying the major developmental
theories – and one that sees the self as an ongoing process – the basic concept
of self-categorisation (for recent summaries see Onorato & Turner, 2001;
Turner, 1999a). In the final section we will suggest ways in which a more fluid
136 Self-categorisation approach to gender
and collective conceptualisation of self might address some of the short-
comings and add to the rich insights already gained through the classic
approaches to gender development.

THE DEVELOPMENTAL SEQUENCE

The importance of the category labels

Anyone who has attempted to buy a congratulatory card for the birth of a
child will be aware of the necessity of choosing between one that is pink and
one that is blue, between one that trumpets “It’s a BOY!” or “Congratula-
tions on the birth of your baby girl”. Classic developmental studies show us
that the importance adults place on labelling a newborn as male or female is
reflected in their tendency to interpret babies’ behaviour differently (Condry
& Condry, 1976) and to treat babies in very different ways (Lewis, 1972),
depending on their sex.
Whether or not they are reflecting adult preoccupations, infants themselves
appear to be aware of the male/female dichotomy (Fagot & Leinbach, 1994;
Leinbach & Fagot, 1993) and their preverbal awareness of categorical sex
differences is expressed as soon as they can speak. By the time they are 2 years
old, children are beginning to apply sex category labels correctly, to them-
selves and to other people (Fagot & Leinbach, 1989, 1993; Thompson, 1975),
and by the time they are in primary school, boys and girls begin to practise
sex segregation, eschewing voluntary contact with the other sex and engaging
in enthusiastic derogation of it (Maccoby, 1988, 1990; Maccoby & Jacklin,
1987; Serbin, Sprafkin, Elmian, & Doyle, 1984; Thorne, 1993; see also
Powlishta, Chapter 4). As one 11-year-old girl expressed it, “It’s like the boys
and girls are on different sides” (Thorne, 1993, p.63).
The primary school years, the period during which the sexes are most
overtly “at war”, is also the period during which they cling most determinedly
to sex stereotypes for themselves and others. Preschoolers may know which
category of objects, behaviours, and characteristics are “girl things” and
which are “boy things”, but they are not overly concerned if they or another
child engages in cross-gender behaviour. On the other side of the “war zone”,
by mid adolescence girls and boys begin to use criteria such as athletic
prowess, artistic, social, and academic interests, and race and ethnicity as the
basis of their social groupings (Brown, Mounts, Lamborn, & Steinberg, 1993;
Eckert, 1989; Schofield, 1981; see also the section on age/developmental
differences in Powlishta, Chapter 4) and, with this association of self with
other social categories, adolescents adopt a more flexible attitude to gender
stereotypes (Katz & Ksansnak, 1994; Welch-Ross & Schmidt, 1996). It is the
content of the stereotypes with which primary school children so strongly
align themselves to which we will now turn our attention.
 David, Grace, Ryan 137
Stereotype content
What do we believe it means to be male or female? Williams and Best (1990)
found that, nearly universally, males are seen as adventurous, forceful, and
independent, and females as sentimental, submissive, and superstitious. In
the more than 30 countries studied, males were characterised as more intelli-
gent, stronger, and more active than females or, to put it in the words of an
Australian children’s playground chant, “Stupid, stupid gi-irls, stupid, stupid
gi-irls! Girls are weak (chuck ’em in the creek). Boys are strong (like King
Kong)”.
In one of the classic studies to which we have already referred, parents
showed evidence of fostering a reactive/proactive distinction in newborn
children, gently constraining and talking quietly to their infant daughters,
and encouraging alertness and exploratory activity in their sons (Lewis,
1972). Caretakers also tend to be more responsive to boys when they are
assertive and girls when they are communicative (Fagot, Hagan, Leinbach, &
Knonsberg, 1985). The reactive/proactive distinction is further fostered
through the toys and the environment that adults provide for children
(Pomerleau, Bolduc, Malcuit, & Cossette, 1990; Reingold & Cook, 1975.
By school age, children have learnt the stereotype-consistent ways to
behave. Girls’ groups are characterised by games in which there is little
evidence of gross motor activity and an emphasis on “grown-up” social
niceties such as turn-taking and making sure everyone is having a good time.
Boys’ groups, on the other hand, are characterised by rough, often aggressive,
hierarchically organised physical activity (Maccoby, 1988, 1990; Maccoby &
Jacklin, 1987; Serbin et al., 1984; Thorne, 1993).
While both boys and girls very quickly learn how each sex is supposed to
play, dress, and behave, and while both sexes usually act in ways that are
consistent with their gender, this is much more prescriptive for boys than for
girls. More than girls, boys are both evaluated negatively (Feinman, 1981),
and sanctioned severely (Fagot, 1977) for cross-sex behaviour, so it is not
surprising that they are more rigid in their adherence to gender-stereotypical
behaviour for themselves and more damning of others who fail to observe
the “rules” (Jackson & Tein, 1998; Signorella, Bigler, & Liben, 1993). The
differential rigidity in gender stereotypes is at its height, as we might
expect, during the gender wars of primary school, but continues to a relative
extent throughout the lifespan (as shown, for example, by Caldera, Huston,
& O’Brien, 1989; Delk, Madden, Livingston, & Ryan, 1986; Rubin,
Provenanzo, & Luria, 1974).
Why is it more important for males than for females to adhere to gender-
appropriate norms? Why is it worse to be a “sissy” than a “tomboy”? Why is
the most damning judgment that can be made of a boy is that he is like a girl?
A recent United Nations publication provides a compelling answer in noting
that . . . “women constitute half the world’s population, perform nearly
two-thirds of its work hours, receive one-tenth of the world’s income, and
138 Self-categorisation approach to gender
own less than one-hundredth of the world’s property” (1991, p. 3). It is
understandable that wealthy landlords should not want to be mistaken for
indigent tenants!
Of course, this is not to say that children are aware of the statistics of
sexual inequality; indeed, few adults are fully cognisant of them. Children do
observe, however, that their fathers’ jobs are more valued than their mothers’
jobs (Catalyst, 1999; Gutek & Larwood, 1989), that females are more likely
than males to be interrupted when they are speaking (Blakemore, Larue, &
Oljenic, 1979 cited in Bussey, 1986), and from the hours children typically
spend in front of television (up to one third of their waking hours according
to Nielsen Media Research, 1989) they will observe that males have more
authority and are more likely to be mature, materially well-off, professional,
competent, and powerful than females (Signorielli & Bacue, 1999; see also
Bussey & Bandura, 1999).
The importance for boys of avoiding any taint of femininity was illustrated
in the first study of a now-classic report (Bussey & Bandura, 1984) in which
children were shown a video of a team of men and a team of women playing
a game in which each team performed a different sequence of ritualised
actions. When invited to play the game themselves, while most girls copied the
women’s actions, a substantial number copied the men’s actions. Boys copied
the men’s actions almost exclusively. That children are sensitive to the gender
power differential became apparent in a second study, in which the authors
preceded the game video with one that showed either the men’s team or the
women’s team owning and having control over all the toys. When men were
seen to have power there was little change in children’s modelling of the
game, but when women were seen to have power there was a significant
increase in boy’s modelling of the women’s actions. Thus, it would appear
that even more important as a cue for gender-appropriate behaviour than
physical appearance (which we know to be extremely important to children)
was the display of power, which children saw as the province of males.
There is also emerging evidence (David, 2003) that boys are aware of
another abstract element of the sexual status quo: The idea that “people” are
male. When asked to select, from pictures of items such as guns and dolls,
those they would show to a Martian to describe human beings, girls chose a
mixture of female and male items, while boys chose almost exclusively male
items.
In sum, girls and boys are perceived as being different and are treated
differently from birth, in ways that encourage expressive reactivity in girls and
competent proactivity in boys. Children tend to conform to their assigned
roles, with a sharp increase in conformity when they begin primary school
and a decrease when, with adolescence, they begin to form friendship groups
based on criteria other than sex. Throughout, boys are more strongly con-
fined than girls by gender stereotypes, since it is vitally important that they
learn not only to be boys, but not to be girls. It is to the theorised form this
learning takes that we will now direct our attention.
 David, Grace, Ryan 139
THEORIES OF GENDER DEVELOPMENT

From the preceding section it will be clear that there is a strong corres-
pondence between the gender stereotypes that children learn and apply to
themselves and social reality. Thus classic psychoanalytic accounts of gender
identity development are less than satisfactory in that they ignore social
factors and place all the emphasis on intra-individual factors such as the
resolution of infant sexual attraction to the opposite-sex parent (Freud, 1925/
1989) or the consequences of identification with the primary care-giver
(Chodorow, 1978, 1979). The only way social reality can be accounted for
within the framework of these theories is to posit that it is a reflection of
fundamentally and irrevocably different male and female personalities. As
such, the status quo cannot and will not change. This is not only pessimistic,
but flies in the face of sociological and historical evidence. (For a fuller cri-
tique of psychoanalytic theories, see Tavris, 1994, and for a discussion of the
social consequences of individualistic theories, see Turner, 1999b).
Looking at gender identity as taking the opposite direction from that
proposed by psychoanalytic theories – that is, seeing it as something which
goes from the outside to the inside, from society to the individual – is social
learning theory (Bandura, 1969, 1986; Bussey & Bandura, 1999; Mischel,
1966, 1993). An extension and modification of the traditional learning theory
postulated by B. F. Skinner, social learning theory proposes that gender
knowledge is gained from observation of many models of both sexes and
extrapolation of within-category similarities and between-category differ-
ences. Thus both girls and boys learn what is appropriate for both sexes. The
decision to adopt one or the other set of behaviours is also determined by
observation, in this case, of the rewards or punishments associated with it for
different protagonists. Children observe, for example, that females are
rewarded for being nurturant and taking care of their appearance, and disap-
proved of when they act too assertively, while male displays of aggression
meet with approval and male displays of nurturance or concern for appear-
ance are met with at best, derision and at worst, violent abuse. Thus, girls
model their behaviour such as dress, activity choice, and manner on other
girls and women, while boys model their behaviour on boys and men.
Gender identity fits towards the end of the process outlined by social
learning theory: Children model their behaviour on similar others whom
they observe being rewarded for the behaviour, then observe their own
behaviour and conclude they are doing girl or boy things, therefore they must
be a girl or a boy. Cognitive-developmental theory (Kohlberg, 1966) turns this
around and posits that children must first know that they are a boy or a girl,
that is, they must achieve gender constancy, and will then seek out what
society defines as boy-appropriate or girl-appropriate behaviour because
consistency between identity and behaviour is rewarding in itself. Thus,
while both theories explain how the social reality of sex differences is
internalised, social learning theory proposes that society socialises children,
140 Self-categorisation approach to gender
while cognitive-developmental theory proposes that children actively socialise
themselves.
Gender schema theory simply proposes that children organise information
about gender in a specific gender schema – “. . . a cognitive structure, a net-
work of associations that organises and guides an individual’s perceptions”
(Bem, 1981, p. 355). Some children are believed to have more highly developed
schemas than others – that is, they are gender-schematic – and they will be
more likely to understand the world in gendered terms, to remember gender-
consistent information better than gender-inconsistent information, and even
to distort memory to make it fit the gender stereotypes contained in the
schema (Stangor & Ruble, 1987). Their gender schema is incorporated by
children into a broad, more inclusive self-schema.
From the perspective of the authors of this chapter, gender schema theory
stands out by virtue of its explicitly stated concept of the self as a structure,
and of people as being fixed types as a result of the nature of the structure.
We will address the problems we see with such a static concept of self at the
end of the following section, but first will address the more general issues
raised by the dominant theories of gender.

MATCHING THE THEORIES AND THE EVIDENCE

Linda Brannon (2002) states that, “Each of the theories of gender develop-
ment presents an orderly pattern of development, but the research shows a
complex pattern with many components that do not necessarily match the
theories. That is, none of the theories is able to explain all the data from
research on gender development” (p. 146).
The insurmountable problem for cognitive-developmental theory is its
predication of gender constancy as the initiator of children’s search for
gender-consistent behaviour. A substantial number of studies have shown
that gender constancy develops later than the appearance of many other
components of gender knowledge (e.g., Bussey & Bandura, 1984; Grace &
David, 2001, 2003): Children typically do not reach gender constancy until
they begin primary school, but by this time they are well able to correctly
identify males and females and to sort objects into gender-appropriate
categories, and they have a very well-developed sense of what is the “right”
way for a girl or a boy to behave.
However, the very thing that is a problem for the cognitive-developmental
approach underlies its main strength: The articulation of a strong motive for
children to seek gendered behaviour – the implicitly rewarding fit between
self-category labels and behaviour. In contrast to this, social learning theory
has been criticised for portraying children as passive objects of societal con-
ditioning (e.g., Bem, 1985). Children show signs of more active pursuit and
enthusiastic embracing of gender-consistent behaviour than social learning
would predict. Moreover, rather than a smooth, gradual accumulation of
 David, Grace, Ryan 141
knowledge, children show distinct patterning of knowledge acquisition, for
example being able to make category distinctions before fleshing out their
content knowledge.
On the “plus” side for social learning theory is the fact that it can account
for the sex differences in gender development, with boys cleaving more
strongly than girls to strict gender-appropriate guidelines. Children’s under-
standing that power is a male prerogative (Bussey & Bandura, 1984) explains
why boys should be so determined not to be mistaken for a girl.
In the eyes of the authors of the current chapter, more important than any
of the above problems is the fact that none of the dominant theories provide
a rationale for the virulence of the primary school separatist “war”. Maccoby
(1990) suggests that children may keep to same-sex play groups because the
girls find the high-energy, rough-and-tumble of boys’ play aversive, and are
put off by the fact that they can exert no influence in mixed-sex groups. There
is ample evidence to back up Maccoby’s claim, but while such a benign
explanation accounts for separatism, it does not explain the enthusiastic
derogation of the opposite sex, such dramatic assertions as that “boys have
cooties” or “girls have girl germs” (Thorne, 1993; for a social identity
explanation, see Powlishta, Chapter 4).
We will be proposing that self-categorisation theory provides a com-
pelling account of this, and other aspects of gender development that are
only poorly explained by the dominant theories. Before attempting this,
however, it is necessary to address the fundamental difference between
self-categorisation and the other theories – its conceptualisation of the self.

THE SELF AS A PROCESS

Gender schema theory explicitly defines the self as a cognitive structure.

While neither social learning nor cognitive-developmental theory make such
a concrete assertion, implicit in both is the idea that the gender knowledge
children are gaining accumulates in some unitary, stable construct that con-
stitutes their gender identity, part of a larger, stable construct that is the self.
In contrast to this, self-categorisation theory conceives of the self as a
labile, context-dependent process (Onorato & Turner, 2001; Turner, 1985,
1988, 1999a, 1999b). The self is whatever an individual means when, at any
time, they call themselves or think of themselves as “I”, “me”, “us”, or “we”.
The primary determinant of the meaning is the social comparative context.
For example, when, in conversation with her sister, a woman says “I”, this
may mean “the person you grew up with, the one who lent you pocket money,
stole your Barbie doll, taught you to use make-up”. If the same woman uses
the first-person singular in conversation with a work colleague, “I” will mean
something very different, perhaps “the person whose office is next to yours,
whose paper-clips you are always borrowing, who is going to get promoted
before you”.
142 Self-categorisation approach to gender
Thus, even the individual self is labile and socially determined. The mean-
ing is determined both by the social context and the reservoir of memories
and knowledge that will inform the content. This knowledge is what many
theorists believe is the self. When contemporary theorists do make allowance
for some fluidity of self-awareness, they see the “working self” (Markus &
Wurf, 1987) as part of a system composed of many situation-specific selves:
the relational self (Markus & Kitayama, 1991), the remembered self (Neisser
& Jopling, 1997), even the false self (Harter, 1997, 1998). Self-categorisation
theory sees this as lacking in parsimony. Since social contexts are potentially
infinite in number, so are the selves that will be determined by them. The
specific knowledge one brings to bear in a particular circumstance may never
have been part of the self-concept before and may never be again. It seems
wasteful, then, that having been part of the self, the knowledge should be
stored on the “self” side of some cognitive barrier. Surely it is more parsi-
monious to see the self as being only what is brought into consciousness at
any particular time, and to propose that in bringing to consciousness a
current “I” or “me”, one has access to all one’s knowledge, memories, and
understandings, not only those contained in a hypothetical “self” enclosure
or system?
If the individual self is a social process, the collective self is much more so.
When thinking or speaking of oneself as a woman or a man, an Australian,
or an American, part of the meaning of “us” or “we” is gained from, and
shared by, the other “selves” included in the first-person plural pronoun. To
know what “us girls” means, a child must know what is typical or normative
for girls – what “we” have in common that distinguishes us from “them”. As
outlined earlier in the chapter, social learning theorists found that gender
knowledge (as distinct from behaviour) is gained through observation of
males and females, and the making of judgments of within-sex similarity and
between-sex difference (Perry & Bussey, 1979). In the classic modelling study
described previously (Bussey & Bandura, 1984), it was observed that children
did not copy the behaviour of a single same-sex adult, but waited until they
had seen most of the team and could conclude that the actions were not
idiosyncratic, but rather the normative way for males or females to behave in
the circumstances. Thus a parent, no matter how loving or loved, cannot be a
model for appropriate gender behaviour, unless the child’s exposure to the
wider world (for example, through friendship groups and the media) suggests
that the parent is a representative or prototypical male or female (see Harris,
1995, for an extended review and discussion of peer group versus parental
influence).
Importantly, a child’s understanding of what it means to be a girl or a boy
will be what they have worked out is typical, based on the exemplars to which
they have had exposure. Since the number and variety of the exemplars will
increase with age, a child’s concept of “typical” will also change, presumably
in the direction of becoming less prescriptive. Thus a self-categorisation con-
ception of self is consistent with the findings that some children’s gender
 David, Grace, Ryan 143
stereotypes are more strongly influenced by their parents than others, and
that children’s gender stereotypes become more flexible with age.
While knowledge of the nature of other self-category members is one
determinant of the meaning of a particular social identity, knowledge of
the out-group is also necessary, since the meaning of “us” is as context-
dependent as the meaning of “I”. For example, “us girls” in a girl/woman
context may mean “the ones who are noisy and boisterous”, but in a “girl/
boy” context may have the opposite meaning of “the ones who are quiet and
mannerly”.
Recent and ongoing studies at The Australian National University have
shown the flexibility of the self-concept. Onorato and Turner (1996, 1997)
required participants to complete a screening questionnaire that identified
them as “independent schematics” or “dependent schematics” (after Markus,
1977). Some men and some women identified themselves as dependent
schematics and some as independent schematics. In single-sex groups the
participants were then required to discuss the extent to which the traits of
tact, cautiousness, and dependency characterised women as distinct from
men (women’s groups) or the extent to which the traits of independence,
dominance, and aggressiveness characterised men as distinct from women
(men’s groups). Participants then described themselves on a social-self rating
task, and women, including those who had rated “me” as independent in the
first task, indicated that “we” were dependent, while men, including those
who had rated “me” as dependent, indicated that “we” were independent.
Critics could perhaps claim that Onorato and Turner’s results do not
necessarily illustrate flexibility of the self-concept, merely that people have an
individual self and a number of collective selves that may differ from each
other in content. The same cannot be said of Ryan and David’s (2002) study,
which tested the proposal that men have separate self-construals and women
connected self-construals (after Cross & Madson, 1997). All participants
completed a self-description questionnaire in which they indicated the extent
to which they perceived their individual selves as connected to, or separate
from, others. This was preceded by a supposedly different study in which they
were required to think about either groups to which they belonged (the
in-group condition), groups to which they did not belong (the out-group
condition), or about gender. When gender had been made salient, partici-
pants responded to the self questionnaire in the gender-stereotypical way,
with women indicating that they were connected, and men that they were
separate individuals. When self-description was preceded by thinking about
groups to which they belonged, both female and male participants indicated
that they were connected individuals. When self-description was preceded by
thinking about groups to which they did not belong, both female and male
participants indicated that they were separate individuals. As well as showing
(a) that the way people experience themselves as individuals is effected by a
feeling of connectedness with in-group others and separation from out-group
others, and (b) that when gender is salient, women experience themselves as
144 Self-categorisation approach to gender
connected to others, and men think of themselves as separate, this study
shows very clearly that the self is not a fixed construct.
This brings us to consideration of the final aspect of self-categorisation to
be discussed here: salience. We have already outlined our understanding of
the self as a context-dependent process. Context plays its part, not in the
simplistic way it has sometimes been seen to (context A calls forth self-
category B, while context X awakens self-category Y), but as part of an
interactive process.
In order to understand this, imagine that a child is contemplating an
unknown group of children. The children have hair of different lengths, from
virtually none to waist-length, but there seems to be a discontinuity at about
the level of the ears: many children have hair lengths ranging from crew-cut,
to ear-length, while others have hair that ranges from shoulder-blade to
waist-length. There is less difference in the crew-cut-to-ear range and the
shoulder-blade-to-waist range than there is between the short and the long,
and the boy observing the group probably perceives even greater within-
category similarity and between-category difference than actually exists (see
Tajfel & Wilkes, 1963). Thus, the group meets the good comparative fit cri-
terion for being perceived as two distinct categories (i.e., within-category
similarity and between-category difference, Oakes, 1987; for the metacontrast
principle, see Turner, 1985). Because the boy’s father is always telling him to
“act like a man”, because his school teacher refers to her class, not as “chil-
dren”, but as “boys and girls”, because his group of friends at school are
currently mounting a terror campaign against a “sissy”, and because his own
male identity is very important to him (see Doosje & Ellemers, 1997), the
gender categories have high relative accessibility for him (he is “perceiver
ready” to call the categories “boys and girls”, see Oakes, 1987). One more
condition must be met before gender will be salient: The two categories must
have good normative (Oakes, 1987), as well as the good comparative fit they
have already displayed. In other words, the categories must display the
qualities that are typically associated with them. Thus, if most of the “short-
hairs” are wearing skirts and most of the “long-hairs” are wearing football
boots, gender will have poor normative fit, despite good comparative fit and
high relative accessibility. In this context, in other words, gender will not be
salient. The categories will not be understood to be boys and girls, and the
observer’s own male identity will thus not be salient (he may think of himself
as a “normal” child in comparison to the strange ones he is watching).
From the earlier part of the chapter we know that such poor normative fit
is highly unlikely. Children, particularly between the ages of 4 years and early
adolescence, religiously follow the dictates of gender appropriateness. “Short-
hairs” are unlikely to wear skirts, or “long-hairs” football boots. Thus, in real
life, children’s behaviour has excellent normative fit with the gender categories.
It also has good comparative fit because of the rigidity with which they
adhere to same-sex modes of being in the world, avoiding at all costs looking
or acting like the opposite sex. Also, because their world is one in which,
 David, Grace, Ryan 145
rather than being called by their own name, they are so often referred to as
one of the boys or the girls who can “tip-toe out quietly”, “line up by the
door”, “go out to little lunch”, and so on, the gender categories are highly
accessible. In sum, gender is very salient for children.
Of course, we already know this as an empirical fact. Indeed, it is difficult
to imagine any other stage in life when a single social category is as salient as
gender is for primary school children. This does not mean, however, that we
can assume that their gender is the only social category with which children
can, or do, identify. We cannot take the lazy option of saying that gender
salience is “chronic” for children (thus falling prey to speaking of gender
differences as if they were part of the “personality”, and taking on board all
the problems associated with seeing the self as a fixed structure). One of the
studies in our current research programme (David, in preparation) illustrates
very clearly that, while gender is salient for children, it is not always so. The
participants (aged 6 to 7 years old) allotted picture stimuli to the “girls’ box”
or the “boys’ box” (gender condition) or to the girls’ box, the boys’ box or the
“grown-ups’ box” (age condition). Other than by the provision of the boxes,
no attempt was made to make them conscious of either the dimension of age
or of gender. In the gender condition, both boys and girls sorted the stimuli
along gender-stereotypical lines (dolls for the girls, guns for the boys, etc.). In
the age condition, although boys’ stereotyping scores were significantly
higher than girls’, both girls and boys had significantly lower gender stereo-
typing scores for their “boy” and “girl” boxes than did children in the gender
condition. Identification measures supported what the stereotyping scores
suggested: Children in the gender condition were thinking of themselves
as girls and boys, but in the age condition, they were primarily aware of
themselves as “children”.
It is difficult to see how any of the dominant theories of gender develop-
ment, embracing the fixed model of the self as they do, could explain this
data. Self-categorisation theory, with its detailed analysis of salience and fluid
understanding of the self, as well as explaining this data, could perhaps
address some of the other questions of complexity that, as Brannon (2002)
pointed out, the major theories have thus far been unable to answer. By the
same token, the predictably high salience of gender for children makes gender
identity a potentially rich vein of inquiry for self-categorisation researchers.
In the following section we direct our attention to both of these assertions.

SELF-CATEGORISATION, THE DOMINANT THEORIES,

AND GENDER DEVELOPMENT

Adopting the self-categorisation concept of the self as a process provides a

different perspective on the dominant theories of gender development. When
social learning theorists propose that children observe males and females
and make evaluations of intracategorical similarity and intercategorical
146 Self-categorisation approach to gender
difference, we would propose that children are assessing the comparative fit of
the gender categories. When children accumulate knowledge about what is
“appropriate” for each category we would see them as laying down a store of
information from which they will make decisions about normative fit, and on
which they will base their own behaviour when gender is salient. The fact
that, as social learning theorists compellingly point out, so much of what
children observe tells them that gender matters (because it determines what
they can wear, how long their hair should be, what and who they can play
with, how they should talk, what school subjects they are good at, what jobs
they do, what sports they play), we would see as meaning that gender will be
highly accessible for children.
Framed this way, the social learning account of gender development
no longer appears to characterise children as passive (Bem’s, 1985, criticism).
We see them as active participants whose current knowledge and level of
cognitive ability determine what they mean when, at any time, they refer to
themselves as a boy or a girl, which in turn determines how they will behave
in that context.
While (with most other contemporary theorists) we reject Piaget’s pro-
posal of 8–9 years as the age at which children can competently categorise, we
do accept that children are not born with fully developed cognitive skills and
capacities, and we cannot ignore the findings that they are able to deal with
simple and concrete processes before complex and abstract ones. The calcula-
tion of comparative fit would have to qualify as complex, and since it is a
necessary part of the process of social identification, we would speculate that
early use of gender labels and preference for gender-appropriate toys are not
driven by the social self-process as outlined in the previous section. In their
early use of “girl” or “boy” to describe themselves, we would suggest that
children are simply applying a personal label, like “Jill” or “John”, that it is
part of the knowledge they have available to use in the personal self-process.
When growing cognitive sophistication enables the perception of social cat-
egories, of similarities within and differences between the categories, and of
identification with one of the categories, the same knowledge can be used to
provide meaning in the social self-process. Because self-categorisation posits
that the self-process draws on all the memories, knowledge, and skills that are
available to an individual, rather than seeing these resources as the self,
changes in children’s cognitive capacities and content are not problematic.
The problem remains for social learning theory, which proposes that child-
ren’s observation of intrasex similarities and intersex differences forms the
basis of their gender knowledge and that only when they observe themselves
behaving in a way that is consistent with one of the categories do they deduce
that they are boys or girls. Children’s gender-consistent labelling, toy and
clothing preferences, and general behaviour prior to making this deduction
can only be seen in this formulation as a series of passive-conditioned
responses, unrelated to gender identity.
Although, as already stated, we disagree with his underlying rationale and
 David, Grace, Ryan 147
proposed timing, we would endorse Kohlberg’s assertion that children are
not passive and will be motivated to seek identity-consistent behaviour.
Where we differ from Kohlberg is that we believe the meaning of self is
constantly changing, that a less complex understanding, or one that has less
correspondence with social or physical reality, is no less “genuinely” self than
a more complex one, or one that reflects the physical and social world more
closely. Thus we would see Jill’s claim that she is a girl because she is wearing
her new frilly socks as a genuine expression of her gender identity, not merely
a preparation for it.
Of course, we cannot ignore the fact that important and dramatic changes
occur in girls’ and boys’ behaviour when they reach school age. Whereas,
however, for social learning and cognitive-developmental theories, this
change is the gaining of “real” gender identity, self-categorisation theorists
would see it as the beginning of gender as a social identity. Prior to the age of
5 or 6 years, children cannot make complex categorical comparisons, nor are
they typically treated as group members. With developments in their cogni-
tive sophistication, and exposure to a social context that makes gender highly
salient, most boys and girls will, for the first time, experience themselves as
“us”. Being a girl or a boy is no longer simply synonymous with being Jill or
John; its meaning is determined by a perception of sameness with others of
the same sex, and differences from those of the opposite sex.
An important consequence of identifying with a social category is that
“we” are motivated to perceive “us” as being better than “them”, or to use the
terminology of social identity theory (Tajfel, 1972; Turner, 1975, 1978), there
will arise “. . . a need for positive social identity, expressed through a desire to
create, maintain, or enhance the positively-valued distinctiveness of ingroups
compared to outgroups” (Turner & Onorato, 1999, p. 18). The form this
commonly takes is that intragroup similarity and intergroup difference is
perceptually exaggerated in a direction that favours the in-group (for evidence
of this in the original minimal group paradigm see Tajfel, Flament, Billig, &
Bundy, 1971; and for a full discussion see Powlishta, Chapter 4). Thus, when
gender is salient – not when the individual self or a social category other than
gender is salient – school-age children will tend to perceive all boys as the
same, all girls as the same, and boys and girls as fundamentaly different from
each other. Girls see the difference as favouring them, while boys see it as
favouring them.
If we apply the Piagetian idea that children rehearse new cognitive pro-
cesses in a way that can seem obsessive to adults, we would predict that,
armed with new social gender identities, primary school children will practise
intergroup skills, and will devote much effort to creating, maintaining, and
exaggerating ways in which girls/boys are better than boys/girls – as of
course we know they do. Thus both the gender separatism and the intergroup
hostility of childhood are explained as simply an illustration of the social
psychology of intergroup relations, albeit a very clear and dramatic one.
By the time the skills are well learnt and can become a less focal part of the
148 Self-categorisation approach to gender
repertoire, children are beginning to have a broader and more varied social
world, access to dimensions other than gender on which to base social iden-
tities, and have reached puberty, which gives the heterosexual majority strong
motivation to perceive members of the opposite sex as potential partners
rather than as enemies. Thus, gender loses its appearance of “chronic” sali-
ence, it becomes less important to “maximise the difference”, and gender
stereotypes become less prescriptive.
We suggest that there are other important facts that need to be considered
concerning the gender intergroup hostility of childhood. Earlier we drew
attention to the fact that people are motivated to gain collective self-esteem
from intergroup encounters. Overt social competition is not, however, pos-
sible when the status quo assigns one of the categories to an inferior position
(Tajfel, 1978). In this case, members of the low-status group do not employ
the empty rhetoric of “we are better than them”, but employ any one of a
number of individual or collective strategies, choice of which is determined
by their perceptions of the stability and legitimacy of the status differential
and of the permeability of the boundaries between the groups (for a sum-
mary see Tajfel & Turner, 1979). Since it is clear that the status quo places
women in an inferior position to men, it should follow that girls will not
enagage in direct social competition with boys – yet we know that they do,
and are in fact more likely than boys to engage in gender in-group favourit-
ism. To understand this, we need to consider a number of developmental
factors that make girls’ situation different from that of women.
First, girls develop earlier than boys, so for most of the primary school
years they are at least as big as, if not bigger than, members of their gender
out-group (Wong, 2000). Girls are also boys’ equals or, in the case of lan-
guage skills, their betters in school work (Halpern, 1997). Thus when boys
assert that girls are “weak“ and “stupid”, this may be backed up by societal
attitudes, but not by the daily reality of girls’ lives. To use Tajfelian termin-
ology, girls will perceive the status differential as illegitimate. Since most have
been protected from the cruelty of the real world, they will probably also
perceive the status differential as unstable. Girls, therefore, do not initially
perceive themselves as deserving lower status than boys, and respond to the
status quo by direct competition for positive distinctiveness.
With time, however, girls’ perceptions of legitimacy and stability will
change. They will observe, for example, that while most primary teachers are
women, senior teachers are usually men, and that high school teachers of the
“objective” subjects like science and maths are men, leaving the less valued
“expressive” subjects like language, music, and art, to women (American
Association of University Women, 1992). Girls will begin to notice that
teachers pay more attention to boys, and that boys respond to this by domin-
ating the classroom: Calling out, demanding help or attention, and being
praised by teachers, while girls themselves receive praise mainly for being
compliant (Epperson, 1988; Sadker & Sadker, 1986). As young adolescents,
girls will come to realise that being popular is determined by how decorative
 David, Grace, Ryan 149
they are, rather than how intelligent, interesting, friendly, or funny (Suitor &
Reavis, 1995). Thus feisty female children will internalise the social reality of
sexual inequality and become disempowered young women. In Tajfelian
terms, they perceive their status as legitimate and realise that the status
boundaries are stable. They stop fighting (at least most of them do!).
The strategies older females subsequently use to gain positive distinctive-
ness for their female identity represent a field of inquiry largely unexplored by
self-categorisation theorists, but one we predict would be rich in content and
perhaps in theoretical insights. Anecdotally, we believe that some women
identify with the female role as defined by the status quo (feminists call these
“male-identified” women, on the basis that the status quo is male-dominated,
male-defined, and male-serving), and gain self-esteem from being good
mothers, daughters, and wives, while others genuinely identify with males and
gain self-esteem from running a successful business, having professional sta-
tus, and in other ways not being like “stupid women”. For most, we suggest, it
is not as straightforward as either of these options, since their female identity
will not always be salient and, when it is, it will mean different things to them
in different contexts. Thus potential researchers will need to take into account
complex interactions of comparative fit, perceived normative fit, and,
importantly, the myriad factors that contribute to perceiver readiness, which
determine not only when the female social identity is salient but what the
identity means, for any particular woman at any particular time (see
Breinlinger & Kelly, 1994).
We believe that there are also developmental changes for males, perhaps
not as much in the meaning of their gender category (for we have seen that
the male stereotype is comparatively less flexible than the female), but in its
salience. As we have seen in this chapter, being male seems to be more
important for boys than being female is for girls. Evidence suggests, on the
other hand, that while gender is highly salient for women, men are more likely
to think of themselves as “people”, “humans”, or whatever the generic cat-
egory of salience happens to be (Hurtig & Pichevin, 1990). The adult situ-
ation is a logical outcome of a society in which the generic “man” is male
(Spender, 1985). In an elegant illustration of this, Eagly and Kite (1987)
asked participants to describe people of specific nationalities, then men, then
women of those nationalities. Descriptions of men of each nationality
matched descriptions of people of that nationality. Descriptions of women
did not match those of men or people of their nationality, but were similar to
women of other nationalities. In other words, men were characterised as
people (of French, Italian, British, or American nationality), while women
were characterised primarily as women.
Thus we suggest that changes in the salience of gender, as well as reflecting
the different circumstances of individuals’ lives, reflect developmental
changes in adaptation to the social reality of a world in which females and
males are not equal. We therefore do not agree with Powlishta’s conclusion
(Chapter 4) that better gender relations can be engendered by encouraging
150 Self-categorisation approach to gender
children to self-categorise on a dimension other than gender. The self-
categorisations that are relevant to us, we believe, are not simply a matter of
choice but rather a direct consequence of the genuine economic, political,
and social circumstances of our lives.

SUMMARY AND SUGGESTIONS

To summarise gender development from the perspective of self-

categorisation theory, we first emphasise that gender is one aspect of the
flexible, changing, context-dependent self-process. What it means to a child to
be a girl or a boy will change with context, as it does for adults, but the
changes will be more dramatic due to developmental growth in cognitive
skills and knowledge, the latter which children will accumulate through direct
and indirect reinforcement and self-directed observation.
As a reflection of their cognitive limitations, early gender identity will
simply be one aspect of a child’s personal identity: They will learn that Jill/
girl plays with dolls or that John/boy is going to be a truck driver when he
grows up. John/boy will come to understand that he meets with approval
when he does things like his father and brother but is ignored, shunned, or
punished when he does things like his sister or mother.
As growing cognitive sophistication allows children to extrapolate from
their observations that, on gender-relevant dimensions, most females have
certain things in common, most males have certain things in common, and
that there is little common ground between females and males, they will form
cognitive representations of the gender categories. Identification with one of
the categories will come with the recognition that they (as Jill/girl or John/
boy) share not just the name, but also the characteristics of one of the gender
categories. In their new world of school, Jill and John will frequently be
referred to, and required to act, as one of “the girls” or “the boys” rather than
as Jill or John; thus the connection between gender category and self will be
reinforced. Boys will receive validation from other in-group members of their
earlier decision to eschew all things female, and both boys and girls will
enthusiastically practise maximising the difference between the sexes.
As they approach adolescence, children’s social worlds will broaden and
they will begin to specialise in activities that suit their unique skills and
enthusiasms. The discovery of similarities between themselves and others
in the choir, the football team, the debating club, the ballet class, and so on,
will increase the dimensions on which they can base social identities, and
gender will lose its prepotency. Intergroup hostility between girls and boys
will lessen as girls adapt to the adult status quo, and as romantic and sexual
partnerships become a possibility.
For girls, since the world will continue to treat them primarily as female,
their gender will continue to be highly salient. While their masculinity will
always be important to them, boys will become the generic or “default
 David, Grace, Ryan 151
option” adult category members and may even come to deny the revelance of
gender in everyday social interactions.
A substantial body of self-categorisation and social identity research
supports many of the premises on which this speculative account of gender
development are based: the flexibility and context-dependent nature of the
self-process; the role of fit and accessibility in determining the salience of self-
categorisations; the motivation to seek collective self-esteem in intergroup
encounters; the fact that lower-status category members use strategies
other than overt social competition to gain positive distinciveness. Similarly,
developmental research confirms such aspects as children’s growing ability to
form cognitive representations of categories, sex differences in the rigidity of
gender stereotypes and in their application to self, and the broadening social
context of adolescents’ lives.
Many aspects, however, need to be investigated and questions remain to be
answered: Is there support for our speculation that there is a developmental
“switch” in the comparative salience of gender as girls and boys become
women and men?; Are young females less aware (or simply less accepting)
than older females, of the gender status quo?; What strategies do adolescent
girls and older women use to gain positive distinctiveness, and how do
these relate to perception of stability, legitimacy, and permeability of the
group boundaries?; What specific variables feed into “perceiver readiness”
to employ gender self-categorisation?; How does self-categorisation as
“generic” effect perceptions and treatment of nongeneric category members?
Regardless of future answers to these questions, self-categorisation theory
can make a valuable contribution to the field. At the practical theoretical
level, we conceptualise the self (and therefore gender identity) as a process,
and understand the descriptive content of a current identity as drawing on all
one’s cognitive resources. Thus we do not have the problems of both cognitive-
developmental and social learning theories in accounting for gendered
behaviour in infants and preschoolers (for cognitive-developmental theory,
the problem is that there is no motivation for gendered behaviour to exist at
all prior to gender constancy; for social learning theory, the problem is that it
can only be seen as a passive conditioned response). Equally importantly, our
understanding of intergroup behaviour as an outcome of social identity pro-
vides an explanation for the hostility of primary school gender interactions,
and for its cessation in adolescence.
At a more abstract theoretical level, because we do not see the descriptive
content of self-awareness as the self, we are able to make a distinction
between the content and the social-psychological process by which indi-
viduals apply the content to their ever-changing self-awareness and through
which the content of gender knowledge informs their behaviour. Thus, while
we continue to investigate the ways in which males and females adapt and
respond to the status quo, we do not see their responses as being the outcome
of stable differences in the “personalities” or “selves” of males and females.
In fact, we would suggest that commonly found sex differences in behaviour
152 Self-categorisation approach to gender
and personality are common human responses to different, established posi-
tions in the social hierarchy. While not referring to the gender categories,
Reynolds and Turner (2001) express this point when they say that “. . .both
groups are engaged in the same psychology; . . . their respective perspectives
stem from the same categorical process in interaction with intergroup
relations and social structural factors” (p. 173).
In our society, gender is arguably “the first socialization” (Bussey, 1986)
and girls and boys are different. Not only do they characterise themselves in
ways that are diametrically opposed (although some would claim comple-
mentary), but there are also sex differences in the intensity with which they
adhere to the characterisations, and in their readiness to invoke gender iden-
tity. We have presented here a perspective that could logically be extended to
suggest that if some other social category, for example class or race, was the
primary dimension on which a society was divided, it would form the basis
for children’s first social identity and possibly the content of the stereotypes
children learnt to apply to themselves would bear the familiar reactive/
expressive versus proactive/competent characterisations of the stereotypes
based on sexual inequality, as Celia Ridgeway’s (2001) argument suggests
might be the case.
The development of gender identity, we thus suggest, should not be
approached as a field of inquiry with its own unique, psychological processes
and specialist theories, but as a field that offers a rich and complex example
of the way children develop self-awareness as members of the society in
which they live.

REFERENCES

American Association of University Women (1992). The AAUW report: How schools
short-change girls. Washington, DC: American Association of University Women
Education Foundation and National Educational Association.
Banaji, M. R., & Prentice, D. A. (1994). The self in social contexts. Annual Review
of Psychology, 45, 297–332.
Bandura, A. (1969). Principles of behavior modification. Stanford, CA: Stanford
University Press.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.
Bem, S. L. (1981). Gender schema theory: A cognitive account of sex-typing.
Psychological Review, 88, 354–364.
Bem, S. L. (1985). Androgeny and gender schema theory: A conceptual and empirical
integration. In T. B. Sonderegger (Ed.), Nebraska symposium on motivation, 1984:
Psychology and gender (pp. 179–226). Lincoln, NE: University of Nebraska Press.
Blakemore, J. E. O., Larue, A. A., & Oljenik, A. B. (1979). Sex-appropriate toy
preferences and the ability to conceptualize toys as sex-role related. Developmental
Psychology, 15, 339–340.
Brannon, L. (2002). Gender: Psychological perspectives (3rd ed.) Boston, MA: Allyn
& Bacon.
 David, Grace, Ryan 153
Breinlinger, S., & Kelly, C. (1994). Women’s responses to status inequality. Psychology
of Women Quarterly, 18, 1–16.
Brooks, J., & Lewis, M. (1976). Infants’ responses to strangers: Midget, adult and
child. Child Development, 47, 323–332.
Brown, B. B., Mounts, N., Lamborn, S. D., & Steinberg, L. (1993). Parenting practices
and peer group affiliation in adolescence. Child Development, 64, 467–482.
Bussey, K. (1986). The first socialization. In N. Grieve & A. Burns (Eds.), Australian
women: New feminist perspectives. Melbourne, Australia: Oxford University Press.
Bussey, K., & Bandura, A. (1984). Influence of gender constancy and social
power on sex-linked modeling. Journal of Personality and Social Psychology, 47,
1292–1302.
Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and
differentiation. Psychological Review, 106, 676–713.
Caldera, Y. M., Huston, A. C., & O’Brien, M. (1989). Social interactions and play
patterns of parents and toddlers, with feminine, masuline, and neutral toys. Child
Development, 60, 70–76.
Catalyst (1999). Catalyst census of women corporate officers and top earners.
Retrieved July 12, 2001, from http://www.catalystwomen.org/press
Chodorow, N. (1978). The reproduction of mothering: Psychoanalysis and the sociology
of gender. Berkeley, CA: University of California Press.
Chodorow, N. (1979). Feminism and difference: Gender, relation, and difference in
psychoanalytic perspective. Socialist Review, 46, 42–64.
Condry, J., & Condry, S. (1976). Sex differences: A study of the eye of the beholder.
Child Development, 47, 812–818.
Cross, S. E., & Madson, L. (1997). Models of the self: Self-construals and gender.
Psychological Bulletin, 122, 5–37.
David, B. (2003). Children’s gender stereotypes and the salience of gender, age and
“human” identities. Manuscript in preparation.
Delk, J. L., Madden, R. B., Livingston, M., & Ryan, T. T. (1986). Adult perceptions of
the infant as a function of gender labelling and observed gender. Sex Roles, 15,
57–534.
Doosje, B., & Ellemers, N. (1997). Stereotyping under threat: The role of group iden-
tification. In R. Spears, P. J. Oakes, N. Ellemers, & S. A. Haslam (Eds), The social
psychology of stereotyping and group life (pp. 257–272). Oxford: Blackwell.
Eagly, A. H., & Kite, M. E. (1987). Are stereotypes of nationalities applied to both
women and men? Journal of Personality and Social Psychology, 53, 451–462.
Eckert, P. (1989). Jocks and burnouts: Social categories and identity in the high school.
New York: Teachers College Press.
Epperson, S. E. (1988, September 16). Studies link subtle sex bias in schools with
women’s behavior in the workplace. Wall Street Journal, p. 27.
Fagot, B. I. (1977). Consequences of moderate cross-gender behavior in preschool
children. Child Development, 48, 902–907.
Fagot, B. I., & Leinbach, M. D. (1989). The young child’s gender schema: Environ-
mental input, internal organization. Child Development, 60, 663–672.
Fagot, B. I., & Leinbach, M. D. (1993). Gender-role development in young children:
From discrimination to labelling. Developmental Review, 13, 205–224.
Fagot, B. I., & Leinbach, M. D. (1994). Gender-role development in young children.
In M. R. Stevenson (Ed.), Gender roles throughout the lifespan: A multi-disciplinary
perspective (pp. 3–24). Muncie, IN: Ball State University.
154 Self-categorisation approach to gender
Fagot, B. I., Leinbach, M. D., & O’Boyle, C. (1992). Gender labelling, gender
stereotyping, and parent behaviors. Developmental Psychology, 28, 225–230.
Fagot, B. I., Hagan, R., Leinbach, M. D., & Knonsberg, S. (1985). Differential
reactions to assertive and communicative acts of toddler boys and girls. Child
Development, 56, 1499–1505.
Feinman, S. (1981). Why is cross-sex-behavior more appropriate for girls than for
boys? A status characteristic approach. Sex Roles, 10, 445–456.
Fiske, S. T. (1992). Thinking is for doing: Portraits of social cognition from
daguerreotype to laserphoto. Journal of Personality and Social Psychology, 63,
877–889.
Freud, S. (1989). Some psychical consequences of the anatomical distinction between
the sexes. In P. Gay (Ed.), The Freud reader (pp. 670–678). New York: Norton.
(Original work published 1925)
Grace, D., & David, B. (2001, July). Sources of influence in preschoolers’ use of gender
categories. Paper presented at the seventh annual meeting of the Society of
Australasian Social Psychologists, Melbourne, Australia.
Grace, D., & David, B. (2003). Sources of influence in preschoolers’ use of gender
categories. Manuscript in preparation: PhD thesis, Australian National University,
Canberra, Australia.
Gutek, B. A., & Larwood, L. (1989). Introduction: Women’s careers are important
and different. In B. A. Gutuk & L. Larwood (Eds.), Women’s career development
(pp. 7–14). Newbury Park, CA: SAGE Publications.
Halpern, D. F. (1997). Sex differences in intelligence: Implications for education.
American Psychologist, 52, 1091–1102.
Harris, J. R. (1995). Where is the child’s environment?: A group socialization theory
of development. Psychological Bulletin, 102, 458–489.
Harter, S. (1997). The personal self in social context: Barriers to authenticity. In R.
Ashmore & L. Jussim (Eds.), Self and identity: Fundamental issues. New York:
Oxford University Press.
Harter, S. (1998). The development of self-representations. In W. Damon (Series Ed.),
N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3, Social, emotional and
personality development (5th ed.), New York: John Wiley & Sons.
Hirschfeld, L. A. (1993). Discovering social difference: The role of appearance in the
development of racial awareness. Cognitive Psychology, 25, 317–350.
Hurtig, M. C., & Pichevin, M. F. (1990). Salience of the sex category system in person
perception: Contextual variations. Sex Roles, 22, 369–395.
Jackson, D. W., & Tein, J.-Y. (1998). Adolescents’ conceptualization of adult roles:
Relationships with age, gender, work goal, and maternal employment. Sex Roles,
38, 987–1008.
Katz, P. A., & Ksansnak, K. R. (1994). Developmental aspects of gender role
flexibility and traditionality in middle childhood and adolescence. Developmental
Psychology, 30, 272–282.
Kohlberg, L. (1966). A cognitive-developmental analysis of children’s sex-role
concepts and attitudes. In E. Maccoby (Ed.), The development of sex differences
(pp. 52–173). Stanford, CA: Stanford University Press.
Leinbach, M. D., & Fagot, B. I. (1993). Categorical habituation to male and female
faces: Gender schematic processing in infancy. Infant Behavior and Development, 16,
317–332.
Lewis, M. (1972). State as an infant–environment interaction: An analysis of
 David, Grace, Ryan 155
mother–infant interaction as a function of sex. Merril-Palmer Quarterly, 18,
58–69.
Maccoby, E. E. (1988). Gender as a social category. Developmental Psychology, 24,
755–765.
Maccoby, E. E. (1990). Gender and relationships. American Psychologist, 45, 513–520.
Maccoby, E. E., & Jacklin, C. N. (1987). The psychology of sex differences. Stanford,
CA: Stanford University Press.
Markus, H. R. (1977). Self-schemata and processing information about the self.
Journal of Personality and Social Psychology, 35, 63–78.
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for
cognition, emotion and motivation. Psychological Review, 98, 224–253.
Markus, H. R., & Wurf, E. (1987). The dynamic self-concept: A social psychological
perspective. Annual Review of Psychology, 38, 299–337.
Mischel, W. (1966). A social-learning view of sex differences in behavior. In E. E.
Maccoby (Ed.), The development of sex differences (pp. 56–81). Stanford, CA: Stan-
ford University Press.
Mischel, W. (1993). Introduction to personality (5th ed.). Fort Worth, TX: Harcourt
Brace Jovanovich.
Money, J., & Ehrhardt, A. (1972). Man and woman, boy and girl. Baltimore, MD:
Johns Hopkins University Press.
Neisser, U., & Jopling, D. (1997). The conceptual self in context: Culture, experience,
self-understanding. New York: Cambridge University Press.
Nielsen Media Research (1989). ’89 Nielson report of television. Northbrook, IL:
Author.
Oakes, P. J. (1987). The salience of social categories. In J. C. Turner, M. A. Hogg, P. J.
Oakes, S. D. Reicher, & M. S. Wetherell (Eds.), Rediscovering the social group: A
self-categorization theory. Oxford: Blackwell.
Onorato, R. S., & Turner, J. C. (1996, May). Fluidity in the self-concept: A shift from
personal to social identity. Paper presented at the second meeting of the Society of
Australasian Social Psychologists, Canberra, Australia.
Onorato, R. S., & Turner, J. C. (1997, April). Individual differences and social identity:
A study of self-categorization processes in the Markus paradigm. Paper presented at
the third meeting of the Society of Australasian Social Psychologists, Woolongong,
Australia.
Onorato, R. S., & Turner, J. C. (2001). The “I”, the “me” and the “us”: The psycho-
logical group and self-concept maintenance and change. In C. Sedikedes & M.
Brewer (Eds.), Individual self, relational self, collective self. Philadelphia, PA:
Psychology Press.
Perry, D. G., & Bussey, K. (1979). The social learning theory of sex differences:
Imitation is alive and well. Journal of Personality and Social Psychology, 37,
1699–1712.
Pomerleau, A., Bolduc, D., Malcuit, G., & Cossette, L. (1990). Pink or blue:
Environmental stereotypes in the first two years of life. Sex Roles, 22, 359–367.
Reingold, H. L., & Cook, K. V. (1975). The content of boys’ and girls’ rooms as an
index of parents’ behavior. Child Development, 46, 510–513.
Reynolds, K. J., & Turner, J. C. (2001). Prejudice as a group process: The role of social
identity. In M. Augostinos & K. J. Reynolds (Eds.), Understanding prejudice, racism,
and social conflict (pp. 159–178). London: SAGE Publications.
Ridgeway, C. (2001, July). Status and the content of stereotypes. Paper presented at the
156 Self-categorisation approach to gender
seventh annual meeting of the Society of Australasian Social Psychologists,
Melbourne, Australia.
Rubin, J. Z., Provenanzo, F. J., & Luria, Z. (1974). The eye of the beholder: Parents’
views on sex of newborns. American Journal of Orthopsychiatry, 44, 512–519.
Ryan, M., & David, B. (2002). A gendered self or a gendered context: A self-
categorization approach to independence and interdependence. Manuscript submitted
for publication.
Sadker, M., & Sadker, D. (1986, March). Sexism in the classroom: From grade school
to graduate school. Phi Delta Kappan, 512–515.
Schofield, J. W. (1981). Complementary and conflicting identities: Images and inter-
actions in an inter-racial school. In S. R. Asher & J. M. Gottman (Eds.), The
development of children’s friendships (pp. 53–90). Cambridge: Cambridge University
Press.
Serbin, L. A., Sprafkin, C., Elmian, M., & Doyle, A. (1984). The early development of
sex differentiated patterns of social influence. Canadian Journal of Social Science,
14, 350–363.
Signorella, M. L., Bigler, R. L., & Liben, L. S. (1993). Developmental differences in
children’s gender schemata about others: A meta-analytic review. Developmental
Review, 13, 147–183.
Signorielli, N., & Bacue, A. (1999). Recognition and respect: A content analysis of
prime-time television characters across three decades. Sex Roles, 40, 527–544.
Spender, D. (1985). Man made language. London: Routledge & Kegan Paul.
Stangor, C., & Ruble, D. (1987). Development of gender role knowledge and gender
constancy. In L. S. Liben & M. L. Signorella (Eds.), Children’s gender schemata (pp.
5–22). San Francisco: Jossey-Bass.
Stevenson, H. W., & Stevenson, N. G. (1960). Social interaction in an interracial
nursery school. Genetic Psychology Monographs, 61, 41–75.
Suitor, J. J., & Reavis, R. (1995). Football, fast cars, and cheerleading: Adolescent
gender norms. Adolescence, 30, 265–272.
Tajfel, H. (1972). La categorisation sociale. In S. Moscovici (Ed.), Introduction a la
psychologie sociale (pp. 272–302). Paris: Larousse.
Tajfel, H. (1978). The social psychology of minorities (Report No. 38). London:
Minority Rights Group.
Tajfel, H., Flament, C., Billig, M. G., & Bundy, R. F. (1971). Social categorization and
intergroup behaviour. European Journal of Social Psychology, 1, 149–177.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In
W. G. Austin & S. Worchell (Eds.), The social psychology of intergroup relations
(pp. 33–47). Monterey, CA: Brooks/Cole.
Tajfel, H., & Wilkes, A. L. (1963). Classification and quantitative judgment. British
Journal of Psychology, 54, 101–114.
Tavris, C. (1994). Reply to Brown and Gilligan. Feminism and Psychology, 4, 350–352.
Thompson, S. K. (1975). Gender labels and early sex role development. Child
Development, 46, 605–608.
Thorne, B. (1993). Gender play: Girls and boys in school. New Brunswick, NJ: Rutgers
University Press.
Turner, J. C. (1975). Social comparison and social identity: Some prospects for
intergroup behaviour. European Journal of Social Psychology, 5, 5–34.
Turner, J. C. (1978). Social categorization and social discrimination in the minimal
group paradigm. In H. Tajfel (Ed.), Differentiation between social groups: Studies in
 David, Grace, Ryan 157
the social psychology of intergroup relations (pp. 101–140). London: Academic
Press.
Turner, J. C. (1985). Social categorization and the self-concept: A social-cognitive
theory of group behaviour. In E. J. Lawlor (Ed.), Advances in group processes:
Theory and research, Vol. 2 (pp. 77–122). Greenwich, CT: JAI Press.
Turner, J. C. (1988). Comments on Doise’s “Individual and social identities in
intergroup relations”. European Journal of Social Psychology, 18, 113–116.
Turner, J. C. (1999a). Some current issues in research on social identity and
self-categorization theories. In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social
identity: Context, commitment, content (pp. 6–34). Oxford: Basil Blackwell.
Turner, J. C. (1999b, July) The prejudiced personality and social change: A
self-categorization perspective. The Tajfel lecture. European Association of
Experimental Social Psychology, Oxford.
Turner, J. C., & Onorato, R. S. (1999). Social identity, personality and the self-
concept: A self-categorization perspective. In T. Tyler, R. M. Kramer, & O. P. John
(Eds.), The psychology of the social self. Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
United Nations. (1991). The world’s women 1970–1990: Trends and statistics. New
York: United Nations Publications.
Walker-Andrews, A. S., Bahrick, L. E., Raglioni, S. S., & Diaz, I. (1991). Infants’
bimodal perception of gender. Ecological Psychology, 3, 55–75.
Welch-Ross, M. K., & Schmidt, C. R. (1996). Gender-schema development and
children’s constructive story memory: Evidence or a developmental model. Child
Development, 67, 820–835.
Williams, J. E., & Best, D. L. (1990). Measuring sex stereotypes: A multination study
(Rev. ed.) Newbury Park, CA: SAGE Publications.
Wong, D. L. (2000). Essentials of pediatric nursing (6th ed.). St. Louis, MO: Mosby.
6 The development of national
identity and social identity
processes: Do social identity
theory and self-categorisation
theory provide useful heuristic
frameworks for developmental
research?
Martyn Barrett, Evanthia Lyons, and
Arantza del Valle

This chapter is concerned with the development of national identity in

children and adolescents, and the possible utility of social identity theory
(SIT) and self-categorisation theory (SCT) as heuristic frameworks for the
investigation of developmental processes in this domain. It is noteworthy
that, while there has been a large amount of research over the years into the
development of both gender and ethnic identity (see, for example, Aboud,
1988; Bernal & Knight, 1993; Eckes & Trautner, 2000; Phinney &
Rotheram, 1987; Ruble & Martin, 1998), comparatively little research has
been conducted into the development of national identity. In addition, SIT
and SCT have proven to be fertile sources of hypotheses for the study of
gender and ethnic identity development. By contrast, that research which
has been conducted into the development of national identity has tended to
be rather descriptive and atheoretical by comparison. The aim of the pres-
ent chapter is to examine whether SIT and SCT might also provide useful
heuristic frameworks for the investigation of the development of national
identity, by providing a range of hypotheses for empirical investigation in
this domain.
The chapter will proceed in the following way. After a brief overview of
some of the principal phenomena that characterise the development of
national identity, the relevance of both SIT and SCT to these phenomena will
be noted. A number of empirical predictions suggested by these two theories
will then be articulated, and these predictions will be examined, primarily
using data drawn from a large multinational study that investigated the sense
of national identity in 6-, 9-, 12-, and 15-year-old British, Spanish, and
Italian children (Barrett et al., 1997). It will be seen that the data are actually
much more complex than might have been anticipated from the perspectives
of SIT and SCT.
160 Development of national identity
To anticipate the conclusions, it will be argued that while SIT and SCT are
useful for the articulation of some key questions for empirical investigation, it
is an open question as to whether these theories themselves provide satisfac-
tory answers to those questions. No simple solutions or answers will be
offered in this chapter. Instead, the present chapter will caution against
resorting to oversimplistic theoretical solutions when addressing what are, in
fact, extremely complex developmental and theoretical issues.
We begin with a brief overview of some of the principal phenomena that
characterise the development of national identity, in order to provide an
initial orientation to some of the issues in this field, and to illustrate why SIT
and SCT might potentially be useful heuristic frameworks for investigating
children’s development in this domain.

THE PRINCIPAL PHENOMENA THAT CHARACTERISE THE

DEVELOPMENT OF NATIONAL IDENTITY

Existing research into the development of children’s sense of national iden-

tity indicates that, before about 5 years of age, children have little knowledge
of their own country or national group (Piaget & Weil, 1951). However, from
about 5 years of age onwards, children can usually state the name of their
own country, and they also begin to classify themselves as members of their
own national group (Barrett, 1996; Wilson, 1998). During mid childhood,
children’s knowledge of the people who belong to their own national group
expands considerably, and by 10 years of age they are able to describe many
of the stereotypical characteristics that are attributed to the members of their
own national group (Barrett, in press; Lambert & Klineberg, 1967; Piaget &
Weil, 1951).
Studies investigating how children feel about their own country and
national group have produced a mixed set of findings. In some studies, it has
been found that children do not have a systematic preference for their own
country or for members of their own national group until 7 years of age or
even later (e.g., Jaspers, Van de Geer, Tajfel, & Johnson, 1972; Middleton,
Tajfel, & Johnson, 1970; Piaget & Weil, 1951). However, other studies have
found a systematic preference for the child’s own country from at least 5–6
years of age (Bennett, Lyons, Sani, & Barrett, 1998; Lambert & Klineberg,
1967; Tajfel, Jahoda, Nemeth, Campbell, & Johnson, 1970; Wilson, 1998). At
whatever age this preference for the in-group is first acquired, once it has been
established, it typically persists through into adolescence (Lambert &
Klineberg, 1967). The importance that children attribute to their national
identity, as well as their degree of identification with their national group,
also tend to increase between 5–6 and 11–12 years of age (Barrett, 2001, in
press).
As far as knowledge of foreign countries and national groups is concerned,
children’s ability to name other countries is poor before about 5 years of age,
 Barrett, Lyons, del Valle 161
and young children also have great difficulty in understanding the concept of
a foreign country (Piaget & Weil, 1951). Knowledge about other countries
begins to develop from about 5 years of age onwards (Barrett & Farroni,
1996; Bourchier, Barrett, & Lyons, 2002; Jahoda, 1962; Piaget & Weil, 1951).
The growth of geographical knowledge in mid childhood is accompanied
by the acquisition and elaboration of stereotypes of the people who live in
other countries and beliefs about their lifestyles and behavioural habits
(Barrett & Short, 1992; Jahoda, 1962; Lambert & Klineberg, 1967; Piaget &
Weil, 1951).
Children sometimes acquire very strong feelings about particular groups of
foreign people before they have acquired any concrete knowledge about those
groups (Barrett & Short, 1992; Johnson, 1973; Johnson, Middleton, & Tajfel,
1970). Despite the general tendency to prefer their own national group over all
others, children can nevertheless feel very positively about some national out-
groups (Barrett, in press; Barrett & Short, 1992; Johnson et al., 1970; Lambert
& Klineberg, 1967; Middleton et al., 1970). The relative order of preference
for other countries, once it is established (perhaps by 6 years of age), seems to
remain fairly stable and consistent across the remaining childhood years
(Barrett & Short, 1992; Jaspers et al., 1972; Johnson et al., 1970). However,
the overall degree of liking for all national out-groups tends to increase
between 5 and 11 years of age (Barrett & Short, 1992; Lambert & Klineberg,
1967). After 11 years of age, this general increase in positive regard for other
national groups usually levels out (Lambert & Klineberg, 1967).
Thus, the studies that have been conducted in this field indicate that, during
mid childhood, children acquire a preference for their own national group,
do not always exhibit out-group denigration, display an increase in positive
regard for national out-groups, and construct stereotypes of the national in-
group and of salient national out-groups. Furthermore, although national
identity may not be especially important to 5-year-old children, it does
become much more important over the course of mid childhood. Children’s
degree of national identification also increases at this time of life, and
national identity typically remains important throughout the adolescent
years.

THE RELEVANCE OF SIT AND SCT TO THE STUDY OF

NATIONAL IDENTITY DEVELOPMENT

SIT and SCT both postulate that in-group favouritism, out-group prejudice,
and the stereotyping of in-groups and out-groups can sometimes occur as
psychological consequences of knowing that one belongs to a particular
social group. These theories posit that these effects may be most pronounced
either when that particular social group membership is subjectively import-
ant to the individual (SIT) or when the prevailing social context renders that
social group membership especially salient to the individual (SCT). There is
162 Development of national identity
thus a prima facie case that these two theories might be able to contribute to
the elucidation of the phenomena of in-group favouritism, prejudice, and
stereotyping as these are exhibited by children and adolescents in relationship
to national groups. In order to examine this possibility in more detail, we first
need to identify some of the principal postulates of these two theories.

Social identity theory

In brief, SIT (Tajfel, 1978; Tajfel & Turner, 1986) is based upon the observa-
tion that individuals belong to many different social groups (e.g., gender,
national, ethnic, occupational, social class, etc.) and that these social group
memberships may sometimes be internalised as part of an individual’s self-
concept. SIT postulates that, when this occurs, individuals strive to obtain a
sense of positive self-worth from these social identities. Consequently, in con-
structing representations of in-groups and relevant out-groups, dimensions
of comparison are chosen that produce more favourable representations of
in-groups than of out-groups, resulting in in-group favouritism or out-group
denigration or both. The positive distinctiveness that is ascribed to the
in-groups produces positive self-esteem.
However, in order for these effects to occur, the individual needs to have
internalised a social group membership as part of his or her self-concept, that
is, the individual must subjectively identify with that category. If an indi-
vidual’s subjective identification with a particular social group is weak or
absent, these effects will not occur. Furthermore, in certain cases (e.g., when
one belongs to a group which has a low social status), it may be difficult to
achieve positive in-group distinctiveness. Under these circumstances, other
strategies (such as changing one’s social group membership, or changing the
existing social order to improve the status of the in-group) may be used
instead to try to achieve a more favourable in-group representation.
SIT also postulates that there is a cognitive processing bias that affects
people’s categorical judgments. This bias attenuates within-category differ-
ences and accentuates between-category differences. Thus, the knowledge
that one belongs to a particular social group produces intragroup homo-
geneity effects in which the variation among the members of in-groups and
out-groups is underestimated, and the differences between members of the in-
group and members of out-groups are overestimated. It is this mechanism
that results in the stereotyping of in-groups and out-groups.
Thus, SIT suggests that in-group favouritism, out-group prejudice, and
stereotyping can occur as psychological consequences of internalising social
group memberships. Over the years, researchers (e.g., Bigler, Brown, & Mar-
kell, 2001; Bigler, Jones, & Lobliner, 1997; Branscombe & Wann, 1994;
Grant, 1992, 1993; Hinkle & Brown, 1990; Kelly, 1988; Mummendey, Klink,
& Brown, 2001; Nesdale & Flesser, 2001; Perreault & Bourhis, 1998;
Powlishta, 1995; Verkuyten, 2001; Yee & Brown, 1992) have drawn a number
of empirical predictions from SIT (although see Turner, 1999, and Haslam,
 Barrett, Lyons, del Valle 163
2001, who have disputed that SIT actually makes such direct predictions, a
point that will be revisited towards the end of this chapter). The predictions
that have been explored by researchers include the following: (1) Representa-
tions of in-groups and out-groups will be based upon dimensions of com-
parison that produce in-group distinctiveness and in-group favouritism. (2)
The strength of identification will correlate either with the positivity of the
in-group evaluation, or with the negativity of out-group evaluations, or with
the positive distinctiveness that is ascribed to the in-group. (3) In-group
favouritism will be a consequence of subjective identification with the
in-group.

Self-categorisation theory
SCT was developed from SIT in order to account for a number of additional
findings that emerged from social-psychological research with adults, espe-
cially the findings that the presence and degree of both in-group bias and
in-group homogeneity are affected by the salience of the relevant social cat-
egorisation in a given setting (Oakes, Haslam, & Turner, 1994; Turner, Hogg,
Oakes, Reicher, & Wetherell, 1987). Elaborating on the basic SIT paradigm,
SCT postulates that individuals have a multiplicity of personal and social
identities, which are organised in the form of a category hierarchy. The level
in the hierarchy at which the self is categorised at any given moment depends
on: the specific social context in which the individual finds him- or herself;
a cognitive process that is driven by the principle of metacontrast (whereby
categorisation occurs at that level in the hierarchy which maximises between-
category differences while minimising within-category differences in the given
context); the fit between the perceiver’s normative beliefs about the particular
categories that are involved and the actual stimuli contained in the current
social situation; and the readiness of the individual to use a particular
categorisation.
When a social identity becomes salient through this mechanism, for
example when the social context contains members of both the in-group and
a relevant out-group, there is a depersonalisation of self-perception (i.e., self-
stereotyping occurs), group behaviour appropriate to the social identity is
elicited, and in-group homogeneity increases. However, when the social con-
text contains only members of the in-group, self-categorisation typically
occurs at a lower level in the hierarchy than that of the in-group, and in-group
homogeneity decreases. In addition, any other factor that enhances the
salience of the social identity for the individual may increase perceptions of
in-group homogeneity. For example, when the in-group is a minority group
perceived as being chronically under threat from a majority out-group, the
salience of the in-group category, the strength of subjective identification,
and in-group homogeneity may be particularly high for those individuals
within intergroup situations.
SCT also proposes that the prevailing comparative context can affect
164 Development of national identity
stereotype content, including the content of the in-group stereotype. The
dimensions used to define the stereotype for a particular social group, as well
as the relative prototypicality of the various group members, can change
according to the comparative context in which the group is being judged.
However, stereotypical variation will be constrained by what the individual
perceiver knows and understands about the particular in-groups and out-
groups that are involved, and about the nature of the intergroup
relationships.
Thus, SCT suggests that in-group bias, the perceived homogeneity of the
in-group, and the content of the in-group stereotype will change as a function
of the prevailing comparative context. However, all of these effects will only
occur if the individual has internalised the relevant social group membership
as part of his or her self-concept. As in the case of SIT, if an individual’s
subjective identification with a group is weak or absent, these various effects
will not occur.
Once again, researchers have drawn a number of empirical predictions
from these various postulates (e.g., Branscombe, Ellemers, Spears, & Doosje,
1999; Brown & Wootton-Millward, 1993; Ellemers, Doosje, Van Knip-
penberg, & Wilke, 1992; Ellemers, Kortekaas, & Ouwerkerk, 1999; Haslam,
Oakes, Turner, & McGarty, 1995; Haslam, Turner, Oakes, McGarty, &
Hayes, 1992; Hopkins & Cable, 2001; Hopkins, Regan, & Abell, 1997; Oakes
et al., 1994; Simon, 1992; Simon & Brown, 1987; Simon & Hamilton, 1994).
These predictions include the following: (1) In-group homogeneity will be
lower in contexts in which only the in-group is present, and higher in contexts
in which relevant out-groups are also present. (2) In-group stereotype content
will change in conjunction with changes in comparative context, with dif-
ferent dimensions being selected depending on the comparison out-groups
that are available in the prevailing context. (3) The strength of identification
will correlate with in-group homogeneity. (4) The strength of identifica-
tion with the in-group may be higher in members of minority groups than in
members of majority groups. (5) In-group homogeneity may be higher
in members of minority groups than in members of majority groups.

THE APPLICATION OF SIT AND SCT TO THE

DEVELOPMENT OF NATIONAL IDENTITY

Thus, a number of predictions have been drawn from both SIT and SCT by
researchers. These predictions may all be tested against the data that are
available from studies into the development of national identity during child-
hood and adolescence. In these studies, data are typically collected from
children at a variety of ages. By examining whether the predicted phenomena
occur at particular ages, it should be possible in principle to obtain some
insight into the developmental process. For example, if the data indicate that
the predicted phenomena do not occur in children of a particular age, then
 Barrett, Lyons, del Valle 165
the full set of social identity processes that are postulated by the relevant
theory as being responsible for the production of those phenomena are either
not yet operative in children of that age, or are being overridden by other
competing factors or processes that are simultaneously operative in those
children. If, however, the data indicate that all of the phenomena predicted by
the relevant theory do indeed occur in children of a given age, then there is
good evidence that the social identity processes postulated by that theory
have begun to operate in children of that age. A more complex outcome
would be one in which some of the predicted phenomena occur, and others
do not, in children of a given age. Such an outcome would be rather more
difficult to interpret theoretically in terms of identifying which particular
processes may be operating in children at that age (although examination of
the phenomena themselves may provide us with some clues). However, for
present purposes, it is sufficient to note that testing these predictions against
the data collected from children at different ages may help to delineate the
sequence in which the various component processes postulated by SIT and/or
SCT begin to operate during the course of development. In the discussion
that follows, each of the eight predictions identified above will be evaluated in
turn against evidence that has been collected on the development of national
identity.

EVALUATION OF THE PREDICTIONS

Representations of in-groups and out-groups will be based upon

dimensions of comparison that produce in-group distinctiveness
and in-group favouritism
As noted earlier in this chapter, existing studies have revealed a somewhat
mixed pattern of findings concerning the age at which a systematic preference
for the national in-group first emerges. These discrepancies may be due to
either of two factors. First, different studies have used different measures of
in-group preference. For example, some studies have measured global affect,
that is, how much the child “likes” the national in-group versus various
comparison out-groups (e.g., Middleton et al., 1970), others have measured
how much the child “likes” particular in-group versus out-group members as
depicted in photographs (e.g., Jaspers et al., 1972), others have asked children
to assign positive and negative adjectives to the in-group and to comparison
out-groups (e.g., Bennett et al., 1998), while others have used open-ended
interviewing (Piaget & Weil, 1951). In addition, different studies have
examined children belonging to different national groups, including Swiss
(Piaget & Weil, 1951), Dutch (Jaspers et al., 1972), English (Middleton et al.,
1970), and British (Bennett et al., 1998) children. It is possible that preference
for the in-group varies from group to group.
Not all of the measures used in these studies are relevant to testing the
prediction. It is most appropriately tested by examining the contents of
166 Development of national identity
children’s descriptions of national groups; SIT predicts that more positive
descriptions will be produced of the national in-group than of comparison
out-groups. This prediction is not appropriately tested using global measures
of how much children like or dislike particular national groups, which
merely assess children’s affect for, and not their representations of, national
groups.
In the multinational study conducted by Barrett et al. (1997), data were
collected from 1700 children aged 6, 9, 12, and 15 years old living in England,
Scotland, Catalonia, Southern Spain, Northern Italy, and Central Italy. Each
of these six subgroups of children was asked to assign descriptive adjectives
both to their own national group and to a number of national out-groups.
The children were instructed to reject adjectives that they considered to be
inapplicable to an individual national group when describing that group.
Because the adjectives consisted of both positive and negative adjectives
(e.g., friendly and unfriendly, clean and dirty), an overall evaluation score
could be computed for each national group based upon the relative propor-
tion of positive and negative adjectives that had been assigned to each group.
The children’s evaluations of their own national in-group were then com-
pared with their evaluations of each individual out-group. The prediction
is that the evaluation of the in-group should be higher than the evaluations
of the out-groups.
The findings are summarised in Table 6.1. This table shows the results
broken down according to whether the children were British (which includes
both the Scottish and the English children together), Spanish (which
includes both the Catalan and the Southern Spanish children together), or
Italian (which includes all of the Italian children together). In the case of the
British and Catalan children, however, it is possible that the British and
Spanish national groups are not the psychologically most relevant groups
with which these children identify, and that these children identify with the
English, Scottish, and Catalan groups instead. Consequently, Table 6.1 also
includes the findings for these three groups of children on their own. The
findings are also broken down according to age.
To start with the youngest children, the table shows that in-group favourit-
ism is clearly exhibited by the Spanish, Italian, English, and Catalan 6-year-
olds; however, in-group favouritism is not exhibited by the 6-year-old Scottish
children, and is not consistently exhibited by the 6-year-old British children
overall. At 9 years of age, in-group favouritism is exhibited by the Spanish
and Catalan children, and by the English children in relationship to all of the
national out-groups except the Scottish out-group. However, at 9 years of
age, the Italian children no longer exhibit in-group favouritism, whereas the
Scottish children are beginning to show some evidence of this phenomenon,
but not consistently so (and while it might be argued that the Italian out-
group is probably not relevant to Scottish children’s self-definitions, it would
be rather more difficult to argue this for the English out-group). At 12 years
of age, the Spanish and Catalan children still consistently display in-group
 Barrett, Lyons, del Valle 167
Table 6.1 Significant differences in the evaluations of the national in-group versus
various national out-groups by 6-, 9-, 12-, and 15-year-old British, Spanish, and
Italian children

British Brit vs. Fren Brit vs. Span Brit vs. Ital Brit vs. Ger

6-yr-olds ns Brit > Span ns Brit > Ger

9-yr-olds ns ns Brit > Ital Brit > Ger
12-yr-olds ns ns ns Brit > Ger
15-yr-olds Brit > Fren ns Ital > Brit Brit > Ger

Spanish Span vs. Fren Span vs. Brit Span vs. Ital Span vs. Ger

6-yr-olds Span > Fren Span > Brit Span > Ital Span > Ger
9-yr-olds Span > Fren Span > Brit Span > Ital Span > Ger
12-yr-olds Span > Fren Span > Brit Span > Ital Span > Ger
15-yr-olds Span > Fren ns Span > Ital Span > Ger

Italian Ital vs. Fren Ital vs. Brit Ital vs. Span Ital vs. Ger

6-yr-olds Ital > Fren Ital > Brit Ital > Span Ital > Ger
9-yr-olds ns Brit > Ital ns ns
12-yr-olds Fren > Ital Brit > Ital Span > Ital ns
15-yr-olds Fren > Ital Brit > Ital Span > Ital ns

English Eng vs. Fren Eng vs. Span Eng vs. Ital Eng vs. Ger Eng vs. Scot

6-yr-olds Eng > Fren Eng > Span Eng > Ital Eng > Ger Eng > Scot
9-yr-olds Eng > Fren Eng > Span Eng > Ital Eng > Ger ns
12-yr-olds ns ns Ital > Eng Eng > Ger Scot > Eng
15-yr-olds Eng > Fren ns Ital > Eng Eng > Ger ns

Scottish Scot vs. Fren Scot vs. Span Scot vs. Ital Scot vs. Ger Scot vs. Eng

6-yr-olds ns ns ns ns ns
9-yr-olds Scot > Fren Scot > Span ns Scot > Ger ns
12-yr-olds ns Scot > Span ns Scot > Ger Scot > Eng
15-yr-olds Scot > Fren Scot > Span ns Scot > Ger Scot > Eng

Catalan Cat vs. Fren Cat vs. Brit Cat vs. Ital Cat vs. Ger

6-yr-olds Cat > Fren Cat > Brit Cat > Ital Cat > Ger
9-yr-olds Cat > Fren Cat > Brit Cat > Ital Cat > Ger
12-yr-olds Cat > Fren Cat > Brit Cat > Ital Cat > Ger
15-yr-olds Cat > Fren Cat > Brit Cat > Ital Cat > Ger

Source: Barrett et al., 1997.

Notes
ns = no significant difference in the evaluations of the two groups.
Brit = British; Fren = French; Span = Spanish; Ital = Italian; Ger = German; Eng = English;
Scot = Scottish; Cat = Catalan.
168 Development of national identity
favouritism, but now the English children no longer do so, while the Scottish
children still only display partial evidence of in-group favouritism. The
12-year-old Italian children actually display a tendency towards in-group
denigration, as they still do at 15 years of age. And at 15 years of age, the
Spanish and Catalan children are still presenting strong evidence of in-group
favouritism (with just a single exception), as are the Scottish children, while
the English 15-year-olds present little consistent evidence of the phenomenon.
Thus, the overall picture that emerges is that in-group favouritism is exhib-
ited to a different extent by different national groups of children at different
ages, and only Spanish and Catalan children behave in a manner consistent
with the prediction of SIT at all ages. Furthermore, different developmental
patterns are exhibited by children who are growing up in different national
contexts. For example, while Italian children exhibit in-group favouritism at 6
years of age but not at later ages, Scottish children do not exhibit this
phenomenon at 6 years of age, but do tend to do so by 15 years of age. Such
differential patterns of development are difficult to explain in terms of there
being particular ages or points in development at which particular identity
processes begin to operate in children.
Barrett et al. (1997) also administered a second measure to assess the child-
ren’s affect towards the various national groups. This took the form of a pair
of questions: “Do you like or dislike X people?” followed by “How much? Do
you like/dislike them a lot or a little?”, with the children’s responses then being
scored on a 5-point scale ranging from “like a lot” to “dislike a lot”. Once
again, the children’s scores for the in-group were compared with the scores
for each of the out-groups. The results are shown in Table 6.2. A very differ-
ent picture emerges from this table. Here, there is clear and consistent evi-
dence of in-group preference in virtually all subgroups of children. However,
while such affective preference for the in-group is interesting as a phenom-
enon in its own right, it is not strictly relevant to testing the prediction made
by SIT. Comparing Table 6.1 with Table 6.2, it is clear that different findings
can be obtained concerning children’s in-group bias depending on the
particular measure used, while Table 6.1 indicates that different findings can
be obtained depending on the particular national population that is being
studied. These observations must be borne in mind when interpreting the
evidence from single nation studies that have utilised single measures.

The strength of identification will correlate either with the

positivity of the in-group evaluation, or with the negativity of
out-group evaluations, or with the positive distinctiveness that is
ascribed to the in-group
In the study by Barrett et al. (1997), two measures were made of the strength
of the children’s national identifications: one that measured the importance
which the children ascribed to their national identity relative to their other
social identities, and one that measured the children’s degree of identification
 Barrett, Lyons, del Valle 169
Table 6.2 Significant differences in the affect expressed towards the national in-group
versus various national out-groups by 6-, 9-, 12-, and 15-year-old British, Spanish,
and Italian children

British Brit vs. Fren Brit vs. Span Brit vs. Ital Brit vs. Ger

6-yr-olds ns ns Brit > Ital Brit > Ger

9-yr-olds Brit > Fren Brit > Span Brit > Ital Brit > Ger
12-yr-olds Brit > Fren Brit > Span Brit > Ital Brit > Ger
15-yr-olds Brit > Fren Brit > Span Brit > Ital Brit > Ger

Spanish Span vs. Fren Span vs. Brit Span vs. Ital Span vs. Ger

6-yr-olds Span > Fren Span > Brit Span > Ital Span > Ger
9-yr-olds Span > Fren Span > Brit Span > Ital Span > Ger
12-yr-olds Span > Fren Span > Brit Span > Ital Span > Ger
15-yr-olds Span > Fren Span > Brit Span > Ital Span > Ger

Italian Ital vs. Fren Ital vs. Brit Ital vs. Span Ital vs. Ger

6-yr-olds Ital > Fren Ital > Brit Ital > Span Ital > Ger
9-yr-olds Ital > Fren Ital > Brit Ital > Span Ital > Ger
12-yr-olds Ital > Fren ns Ital > Span Ital > Ger
15-yr-olds Ital > Fren ns ns Ital > Ger

English Eng vs. Fren Eng vs. Span Eng vs. Ital Eng vs. Ger Eng vs. Scot

6-yr-olds Eng > Fren Eng > Span Eng > Ital Eng > Ger Eng > Scot
9-yr-olds Eng > Fren Eng > Span Eng > Ital Eng > Ger Eng > Scot
12-yr-olds Eng > Fren Eng > Span Eng > Ital Eng > Ger Eng > Scot
15-yr-olds Eng > Fren Eng > Span Eng > Ital Eng > Ger Eng > Scot

Scottish Scot vs. Fren Scot vs. Span Scot vs. Ital Scot vs. Ger Scot vs. Eng

6-yr-olds Scot > Fren ns Scot > Ital Scot > Ger Scot > Eng
9-yr-olds Scot > Fren Scot > Span Scot > Ital Scot > Ger Scot > Eng
12-yr-olds Scot > Fren Scot > Span Scot > Ital Scot > Ger Scot > Eng
15-yr-olds Scot > Fren Scot > Span Scot > Ital Scot > Ger Scot > Eng

Catalan Cat vs. Fren Cat vs. Brit Cat vs. Ital Cat vs. Ger

6-yr-olds Cat > Fren Cat > Brit Cat > Ital Cat > Ger
9-yr-olds Cat > Fren Cat > Brit Cat > Ital Cat > Ger
12-yr-olds Cat > Fren Cat > Brit Cat > Ital Cat > Ger
15-yr-olds Cat > Fren Cat > Brit Cat > Ital Cat > Ger

Source: Barrett et al., 1997.

Notes
ns = no significant difference in the affect expressed towards the two groups.
Brit = British; Fren = French, Span = Spanish; Ital = Italian; Ger = German; Eng = English;
Scot = Scottish; Cat = Catalan.
170 Development of national identity
with their national group (i.e., whether they felt “very British/Spanish/
Italian”, “a little bit British/Spanish/Italian” or “not at all British/
Spanish/Italian”, etc.). The correlations between each of these two scores and
the children’s evaluations of their national in-group and the various national
out-groups were examined. The results are summarised in Table 6.3.
This table provides little support for the prediction that the strength of
national identification will correlate with the evaluation of national in-groups
and out-groups. Only the Spanish children’s evaluations of their own
national in-group produced consistent evidence supporting this prediction. In
addition, the older Catalan children’s evaluations of their national in-group
indicate that a similar correlation emerges during adolescence in these chil-
dren. However, the other significant correlations exhibited by the Catalan
children’s data also indicate that the stronger their sense of national identity,
the more positive their evaluations of some national out-groups at certain
ages, contrary to prediction.
The data were also examined to ascertain whether the strength of national
identification was related to the difference between the evaluation of the in-
group and the evaluations of the out-groups, in other words, to the positive
distinctiveness of the in-group vis-à-vis each individual out-group. Con-
sequently, the correlations between the differences between the various in-
group–out-group evaluations (i.e., the positive distinctiveness of the in-group
over each individual out-group) and the two strengths of national identifica-
tion scores were also examined. The results of these analyses are shown in
Table 6.4. This table reveals that the predicted relationship only applies in the
case of the Spanish children. In addition, there is partial evidence that this
relationship might apply in some Catalan children as far as their Catalan
identity is concerned, but there is no real evidence that it applies in Italian,
English, or Scottish children.
Next, the correlations between the strength of identification measures and
the affect that was expressed towards each individual national group (in
response to the like/dislike questions) were examined. The results are shown
in Table 6.5. First, it is of interest to note that strength of identification with
the in-group does not appear to be related to affect towards national out-
groups. Second, Table 6.5 suggests that the strength of in-group identification
tends to be positively related to the affect expressed towards the national in-
group in the case of the British, Spanish, Italian, and Catalan identities.
However, this relationship was largely absent in the case of the English and
Scottish children as far as their English and Scottish identities were
concerned.
Last, the data were examined to ascertain whether the strength of national
identification was related to the difference between the affect expressed
towards the in-group and the affect expressed towards the out-groups, in
other words, to the affective distinctiveness of the in-group vis-à-vis each
individual out-group. The results are shown in Table 6.6. This table reveals
that the predicted relationship applies consistently only in the case of the
 Barrett, Lyons, del Valle 171
Table 6.3 Correlations between the evaluation of each individual national group
and either the importance of national identity (Imp) or the degree of national
identification (Deg)

British Brit Span Ital Ger Fren

6-yr-olds ns ns ns ns ns
9-yr-olds ns ns ns ns ns
12-yr-olds ns Deg – ns ns ns
15-yr-olds ns ns ns ns ns

Spanish Span Brit Ital Ger Fren

6-yr-olds Imp +, Deg + ns ns ns ns

9-yr-olds Imp +, Deg + Deg – ns ns ns
12-yr-olds Imp +, Deg + Imp –, Deg – Deg – ns ns
15-yr-olds Imp +, Deg + ns ns ns ns

Italian Ital Brit Span Ger Fren

6-yr-olds Imp + ns ns ns ns
9-yr-olds ns ns ns ns ns
12-yr-olds ns ns ns ns ns
15-yr-olds ns ns ns ns ns

English Span Ital Ger Fren Scot

6-yr-olds ns ns ns ns ns ns
9-yr-olds ns ns ns ns ns ns
12-yr-olds ns ns ns ns ns ns
15-yr-olds ns ns ns ns ns ns

Scottish Span Ital Ger Fren Eng

6-yr-olds ns ns ns ns ns ns
9-yr-olds ns ns ns ns ns ns
12-yr-olds ns ns ns ns ns Imp –
15-yr-olds ns ns ns ns ns ns

Catalan Cat Brit Ital Ger Fren

6-yr-olds ns ns ns ns ns
9-yr-olds ns Imp + ns Imp + ns
12-yr-olds Deg + ns ns Imp +, Deg + ns
15-yr-olds Imp +, Deg + ns ns ns Deg +

Source: Barrett et al., 1997.

Notes
ns = no significant correlations; + = positive correlation; – = negative correlation.
Brit = British; Fren = French, Span = Spanish; Ital = Italian; Ger = German; Eng = English;
Scot = Scottish; Cat = Catalan.
172 Development of national identity
Table 6.4 Correlations between the positive distinctiveness of the national in-group (as
measured against each individual out-group) and either the importance of national
identity (Imp) or the degree of national identification (Deg)

British Brit vs. Fren Brit vs. Span Brit vs. Ital Brit vs. Ger

6-yr-olds ns Deg + ns Deg +

9-yr-olds ns Deg + ns ns
12-yr-olds ns Imp – ns ns
15-yr-olds ns ns ns ns

Spanish Span vs. Fren Span vs. Brit Span vs. Ital Span vs. Ger

6-yr-olds Imp + Imp +, Deg + Imp + Imp +

9-yr-olds ns Imp +, Deg + Imp +, Deg + ns
12-yr-olds Imp +, Deg + Imp +, Deg + Imp +, Deg + Imp +, Deg +
15-yr-olds Imp +, Deg + Imp +, Deg + Imp +, Deg + Imp +, Deg +

Italian Ital vs. Fren Ital vs. Brit Ital vs. Span Ital vs. Ger

6-yr-olds ns Imp + ns ns
9-yr-olds ns ns ns ns
12-yr-olds ns ns ns ns
15-yr-olds ns ns ns ns

English Eng vs. Fren Eng vs. Span Eng vs. Ital Eng vs. Ger Eng vs.
Scot

6-yr-olds ns ns ns ns ns
9-yr-olds ns ns ns ns ns
12-yr-olds ns ns ns ns ns
15-yr-olds ns ns ns ns ns

Scottish Scot vs. Fren Scot vs. Span Scot vs. Ital Scot vs. Ger Scot vs.
Eng

6-yr-olds ns ns Imp – ns ns
9-yr-olds ns ns ns ns ns
12-yr-olds ns ns ns ns Imp +
15-yr-olds ns ns ns ns ns

Catalan Cat vs. Fren Cat vs. Brit Cat vs. Ital Cat vs. Ger

6-yr-olds Deg + ns Deg + Deg +

9-yr-olds ns ns ns ns
12-yr-olds Deg + ns ns ns
15-yr-olds ns Imp +, Deg + ns ns

Source: Barrett et al., 1997.

Notes
ns = no significant correlations; + = positive correlation; – = negative correlation.
Brit = British; Fren = French, Span = Spanish; Ital = Italian; Ger = German; Eng = English;
Scot = Scottish; Cat = Catalan.
 Barrett, Lyons, del Valle 173
Table 6.5 Correlations between the affect expressed towards each individual national
group and either the importance of national identity (Imp) or the degree of national
identification (Deg)

British Brit Span Ital Ger Fren

6-yr-olds Imp + ns ns Imp + ns

9-yr-olds Imp +, Deg + ns ns ns ns
12-yr-olds Deg + ns ns ns ns
15-yr-olds Imp +, Deg + ns ns ns ns

Spanish Span Brit Ital Ger Fren

6-yr-olds Imp +, Deg + ns Imp – ns ns

9-yr-olds Imp +, Deg + ns ns ns ns
12-yr-olds Imp +, Deg + Deg – Deg – ns ns
15-yr-olds Imp +, Deg + ns ns ns ns

Italian Ital Brit Span Ger Fren

6-yr-olds Imp +, Deg + ns ns ns ns

9-yr-olds ns Deg + ns ns ns
12-yr-olds Imp + ns ns ns ns
15-yr-olds Deg + ns ns ns ns

English Eng Span Ital Ger Fren Scot

6-yr-olds ns ns ns ns ns ns
9-yr-olds ns ns ns ns ns ns
12-yr-olds Deg + ns ns ns ns ns
15-yr-olds ns ns ns ns ns ns

Scottish Scot Span Ital Ger Fren Eng

6-yr-olds ns ns ns ns ns ns
9-yr-olds ns ns ns ns ns ns
12-yr-olds ns ns ns ns ns ns
15-yr-olds ns ns ns ns ns ns

Catalan Cat Brit Ital Ger Fren

6-yr-olds ns ns ns ns ns
9-yr-olds Deg + ns ns ns ns
12-yr-olds Imp +, Deg + ns ns Deg + Imp +
15-yr-olds Imp +, Deg + ns ns ns Deg +

Source: Barrett et al., 1997.

Notes
ns = no significant correlations; + = positive correlation; – = negative correlation.
Brit = British; Fren = French, Span = Spanish; Ital = Italian; Ger = German; Eng = English;
Scot = Scottish; Cat = Catalan.
174 Development of national identity
Table 6.6 Correlations between the affective distinctiveness of the national in-group
(as measured against each individual out-group) and either the importance of
national identity (Imp) or the degree of national identification (Deg)

British Brit vs. Fren Brit vs. Span Brit vs. Ital Brit vs. Ger

6-yr-olds ns ns Imp + ns
9-yr-olds Imp +, Deg + Deg + ns Imp +, Deg +
12-yr-olds ns ns Imp + ns
15-yr-olds Imp +, Deg + Imp +, Deg + ns Imp +, Deg +

Spanish Span vs. Fren Span vs. Brit Span vs. Ital Span vs. Ger

6-yr-olds Imp +, Deg + Imp +, Deg + Imp +, Deg + Imp +, Deg +

9-yr-olds Imp + Imp +, Deg + Imp +, Deg + Imp +, Deg +
12-yr-olds Imp +, Deg + Imp +, Deg + Imp +, Deg + Imp +, Deg +
15-yr-olds Imp +, Deg + Imp +, Deg + Imp +, Deg + Imp +, Deg +

Italian Ital vs. Fren Ital vs. Brit Ital vs. Span Ital vs. Ger

6-yr-olds ns ns ns ns
9-yr-olds ns ns ns ns
12-yr-olds Imp + Imp + Deg + ns
15-yr-olds ns Deg + Deg + ns

English Eng vs. Fren Eng vs. Span Eng vs. Ital Eng vs. Ger Eng vs.
Scot

6-yr-olds ns ns ns ns ns
9-yr-olds ns ns ns ns ns
12-yr-olds ns ns ns ns Deg +
15-yr-olds ns ns ns ns ns

Scottish Scot vs. Fren Scot vs. Span Scot vs. Ital Scot vs. Ger Scot vs.
Eng

6-yr-olds ns ns ns ns ns
9-yr-olds ns ns ns ns ns
12-yr-olds ns ns ns ns ns
15-yr-olds ns ns ns ns ns

Catalan Cat vs. Fren Cat vs. Brit Cat vs. Ital Cat vs. Ger

6-yr-olds ns ns ns Imp +
9-yr-olds ns ns ns ns
12-yr-olds Deg + Imp +, Deg + Imp +, Deg + ns
15-yr-olds ns Imp +, Deg + ns ns

Source: Barrett et al., 1997.

Notes
ns = no significant correlations; + = positive correlation; – = negative correlation.
Brit = British; Fren = French, Span = Spanish; Ital = Italian; Ger = German; Eng = English;
Scot = Scottish; Cat = Catalan.
 Barrett, Lyons, del Valle 175
Spanish children at all ages. In addition, there is partial evidence that this
relationship applies among some British, older Italian, and older Catalan
children as well, but there is little evidence that it applies in the case of the
English and Scottish children in relationship to their English and Scottish
identities at any age.
Thus, once again, there are differences in findings according to the precise
method that is used to assess in-group bias, and according to which particular
national group is being investigated. Overall, Tables 6.3–6.6 suggest only
limited empirical support for the second theoretical prediction, with different
developmental patterns being exhibited by children who are growing up
within different national contexts.

In-group favouritism will be a consequence of subjective

identification with the in-group
SIT postulates that in-group favouritism occurs as a consequence of subject-
ive identification with the in-group. When an individual internalises a social
group membership as part of his or her self-concept, that individual is motiv-
ated by the need for positive self-esteem to construct more positive represen-
tations of that in-group than of out-groups. Thus, in-group favouritism is a
consequence of subjective identification; if the individual does not identify
with that group, then there is no motivation to evaluate the group more
positively than other groups.
Bennett et al. (1998) examined the British data collected by Barrett et al.
(1997), in order to ascertain whether or not subjective identification is, in fact,
a necessary precondition for in-group favouritism. Using the degree of iden-
tification task, they found that 30% of the British children failed to identify
with being British, and 14% of the children failed to identify with being either
English or Scottish, even though they were all, de facto, both British and
either English or Scottish (these nonidentifying children tended to be the
younger children in the samples). Focusing upon the data of these nonidenti-
fying children, Bennett et al. looked to see whether these children nevertheless
exhibited in-group favouritism. They found that the children who failed to
identify as being British in fact assigned more positive adjectives to the Brit-
ish national group than to any other national group in the adjective evalu-
ation task, and expressed more positive affect towards British people than
towards Italian and German people (although French and Spanish people
were liked just as much as British people). In addition, the children who failed
to identify as being English or Scottish assigned more positive adjectives to
the English or Scottish national group than to any other national group, and
expressed more positive affect towards the English or Scottish national group
than towards any other national groups.
These findings indicate that nonidentifying children may exhibit favourit-
ism towards their own national group, even though they do not categorise
themselves as being members of that group. From a developmental
176 Development of national identity
perspective, this finding is important, as it implies that such favouritism must,
at least in some cases, stem from factors other than subjective identification
with the in-group coupled to the motivation for positive self-esteem.

In-group homogeneity will be lower in contexts in which only the

in-group is present, and higher in contexts in which relevant
out-groups are also present
SCT differs in its emphasis from SIT, insofar as it predicts that the salience of
an individual’s national identity, and therefore the perceived homogeneity of
the in-group, will be higher in contexts in which relevant out-groups are
present than in contexts in which only the in-group is present. It is postulated
that the latter kind of situation encourages intragroup comparisons, which
therefore enhance the salience of lower-level (personal or subgroup) categor-
ies between which the differences are accentuated, thereby increasing the
perceived variability of the in-group. By contrast, in intergroup comparative
contexts that contain both the in-group and a relevant out-group, intergroup
comparisons occur and intragroup differentiations within the in-group
become less salient, with the result that the perceived variability of the
in-group decreases and the salience of the in-group category increases.
In order to examine whether these effects occur in children in relationship
to their national identity, Barrett, Wilson, and Lyons (1999) conducted an
experiment with 5- to 11-year-old English children. The children were asked
to attribute adjectives to their own national group either on its own or in
conjunction with attributing characteristics to one of two national out-
groups, either Americans or Germans. From the attribution of adjectives, a
perceived variability score for each of the target national groups was derived.
This was essentially a measure of how variable each of those national groups
was perceived to be on those descriptive dimensions that the children them-
selves had used for attributing the adjectives to those groups. In addition, the
importance that the children ascribed to their own English national identity
in relationship to their other social identities was assessed immediately after
the administration of the attribution task.
It was found that the younger children attributed less variability to all of
the national groups than the older children. In addition, the younger children
ascribed less importance to their national identity than the older children.
However, judging the in-group in the presence of an American or German
out-group did not reduce the perceived variability of the national in-group
compared with when that in-group was judged on its own. The presence of a
comparative out-group also did not increase the importance that was
ascribed to the national identity relative to other identities (such as gender
and age). In other words, in-group homogeneity was not lower in the non-
comparative context than in the two comparative contexts, and the relative
importance of the national identity was also not lower in the noncomparative
condition than in the comparative conditions.
 Barrett, Lyons, del Valle 177
Interestingly, this study did find that in-group favouritism (as indexed
using an evaluation measure) was exhibited by the English children at all
ages in relationship to both Germans and Americans. It will be recalled
that in-group favouritism was also displayed by the English children in
relationship to Germans in the study by Barrett et al. (1997; see Tables
6.1 and 6.2). However, there were no significant correlations between
the importance of the English national identity and either the in-group
evaluation, the two out-group evaluations, or the positive distinctiveness
of the English in-group vis-à-vis either the American or the German out-
group (which is consistent with the lack of such relationships among
the English children in the study by Barrett et al., 1997; see Tables 6.3
and 6.4).

In-group stereotype content will change in conjunction with

changes in comparative context, with different dimensions being
selected depending on the comparison out-groups that are
available in the prevailing context
In the study by Barrett et al. (1999), the adjectives that the children attributed
to the English in-group in the three different conditions (i.e., evaluating
English people on their own, evaluating English people in relationship to
Germans, and evaluating English people in relationship to Americans) were
also examined to see whether the children attributed different characteristics
to the in-group in the three conditions. It was anticipated that different
characteristics would be attributed to English people when they were being
compared to Americans and to Germans (e.g., it was expected that “friendly”
would be used less frequently in the American context, and “peaceful” would
be used more frequently in the German context, because English children
typically perceive Americans as “friendly” and Germans as “aggressive”:
Barrett, Day, & Morris, 1990; Barrett & Short, 1992; Wilson, Barrett, &
Lyons, 1995). However, there were no effects of comparative context upon the
contents of the children’s in-group stereotype. That is to say, irrespective of
the out-group against which the children were comparing the English, they
attributed very similar adjectives to the English in-group. Thus, 5- to 11-year-
olds do not appear to be sensitive to the comparative context when attri-
buting characteristics to national groups. Similar findings have also been
obtained by Sani, Bennett and Joyner (1999), who also failed to find any
effect of comparative context on the in-group national stereotypes of
Scottish 7- to 11-year-old children.
Thus, national in-group stereotype content does not appear to change
in conjunction with changes in comparative context in 5- to 11-year-old
children, contrary to prediction. Insofar as these effects of comparative con-
text upon national stereotype content clearly do occur in adults aged 18 years
and older (e.g., Haslam et al., 1992, 1995; Hopkins et al., 1997), it would be
useful to ascertain the age at which these effects begin to occur (currently,
178 Development of national identity
no data are available on 12- to 17-year-olds as far as their national identity is
concerned).

The strength of identification will correlate with perceived in-

group homogeneity
SCT postulates that, when an individual has internalised a social group
membership as part of his or her self-concept, then when that social identity
becomes salient to that individual (perhaps as a function of the prevailing
social context), there is a depersonalisation of the perception of self and
others, the perceived homogeneity of the in-group increases, self-stereotyping
occurs, and group behaviour that is appropriate to the social identity is
elicited. This increase in in-group homogeneity is therefore one of the con-
sequences of subjective identification; if an individual does not subjectively
identify with a group, then that particular group membership will not be so
salient for that individual in the intergroup context, and in-group homo-
geneity will not be so pronounced in such an individual compared with an
individual for whom that social group membership is subjectively important.
Thus, SCT predicts that in-group homogeneity will be higher in individuals
who subjectively identify with the group compared with individuals who do
not identify with the group; in other words, in an intergroup situation, the
strength of identification should correlate with in-group homogeneity.
The data from Barrett et al. (1997) were used to test this prediction. Using
the data from the task in which the children were asked to assign descriptive
adjectives to their own national group and to a number of national out-
groups (in which an intergroup frame of reference had been established), a
perceived variability score for each of the target national groups was derived.
This score was a measure of how variable each national group was perceived
to be on those descriptive dimensions that the children themselves had used
for assigning the adjectives to that group; a high score meant that a group was
perceived to be highly variable, while a low score meant that a group was
perceived to be highly homogeneous. In addition, two measures of the child-
ren’s strength of identification with the national in-group were available: the
importance that the children ascribed to their national identity relative to
their other social identities, and the children’s degree of identification with
their national group. The correlations between each of these two identifica-
tion measures and the perceived variability of each individual national group
are summarised in Table 6.7.
This table shows that only the Spanish 6- and 12-year-olds and the Catalan
6- and 15-year-olds provided evidence consistent with the prediction that
there will be a negative correlation between the strength of identification with
the national in-group and the perceived variability of that group. In addition,
the stronger the 6- and 12-year-old Spanish children’s national identification,
the more they perceived the British, Italian, and French national out-groups
as being homogeneous. However, these trends were not exhibited by the
 Barrett, Lyons, del Valle 179
Table 6.7 Correlations between the perceived variability of each individual national
group and either the importance of national identity (Imp) or the degree of national
identification (Deg)

British Brit Span Ital Ger Fren

6-yr-olds Imp + ns ns Deg + ns

9-yr-olds ns ns ns ns ns
12-yr-olds ns Imp + ns ns ns
15-yr-olds ns ns ns ns ns

Spanish Span Brit Ital Ger Fren

6-yr-olds Imp – Imp – Imp – ns Imp –, Deg –

9-yr-olds ns ns Imp – ns ns
12-yr-olds Imp –, Deg – Deg – Deg – ns Imp –
15-yr-olds ns ns ns ns ns

Italian Ital Brit Span Ger Fren

6-yr-olds ns ns ns ns ns
9-yr-olds ns ns ns ns ns
12-yr-olds ns ns ns ns ns
15-yr-olds ns ns Deg – ns ns

English Eng Span Ital Ger Fren Scot

6-yr-olds ns ns ns ns ns ns
9-yr-olds ns ns ns ns ns Deg –
12-yr-olds ns ns ns Imp + ns ns
15-yr-olds ns ns Imp – ns ns ns

Scottish Scot Span Ital Ger Fren Eng

6-yr-olds ns ns ns ns ns ns
9-yr-olds ns ns ns ns ns ns
12-yr-olds ns Deg – Deg – ns ns ns
15-yr-olds ns ns ns ns ns ns

Catalan Cat Brit Ital Ger Fren

6-yr-olds Imp – ns ns ns ns
9-yr-olds ns ns ns ns ns
12-yr-olds ns ns ns ns ns
15-yr-olds Imp – ns ns Imp – ns

Source: Barrett et al., 1997.

Notes
ns = no significant correlations; + = positive correlation; – = negative correlation.
Brit = British; Fren = French, Span = Spanish; Ital = Italian; Ger = German; Eng = English;
Scot = Scottish; Cat = Catalan.
180 Development of national identity
children growing up in other national contexts. Overall, it would appear from
Table 6.7 that the degree of stereotyping of both national in-groups and out-
groups is not consistently related to the strength of national identification in
children aged between 6 and 15 years of age.

The strength of identification with the in-group may be higher in

members of minority groups than in members of majority
groups
The data collected by Barrett et al. (1997) were also examined to ascertain
whether individuals who belong to a minority national group exhibit a higher
level of identification with their in-group than do the members of a majority
national group. The analyses here initially focused upon the data from the
English and Scottish children. Within the British context, the English
national group is both politically and economically more powerful than the
Scottish national group; the English national group is also numerically much
larger. Thus, the Scottish nation is a subordinate group within Britain in
terms of its power, wealth, and size. Under these circumstances, one might
expect Scottish people to display higher levels of identification with their
national in-group than English people (cf. Branscombe et al., 1999; Simon,
1992).
In order to examine this possibility, both the importance of the national
identity and the degree of identification with the national group were com-
pared in the English and Scottish children. The results are summarised in
the upper half of Table 6.8 (first two columns). This shows that only the

Table 6.8 Significant differences in the importance of national identity, the degree
of national identification, and the perceived variability of the national in-group, in
Scottish versus English children, and in Catalan versus Spanish (Andalusian) children

Importance of Degree of national Perceived variability

national identity identification of the national
in-group

English vs. Scottish

6-yr-olds Eng > Scot Eng > Scot ns
9-yr-olds ns Eng > Scot ns
12-yr-olds ns ns Eng > Scot
15-yr-olds Scot > Eng ns Eng > Scot
Catalan vs. Spanisha
6-yr-olds ns Cat > Span Cat > Span
9-yr-olds ns ns Cat > Span
12-yr-olds ns Span > Cat Cat > Span
15-yr-olds ns ns ns

Notes
ns = no significant difference.
a
In Andalusian children.
 Barrett, Lyons, del Valle 181
15-year-old children displayed the expected pattern (and then only on one of
the two measures of the strength of national identification). Develop-
mentally, this outcome does make sense, as it is probable that only the older
children would have acquired sufficient knowledge concerning the political,
economic, and minority status of Scotland for the predicted effects to occur.
However, this explanation does not account for the fact that the opposite
pattern was exhibited by the younger children (where the English children
actually exhibited a stronger sense of national identification than the Scottish
children).
Similar analyses were also conducted on the data collected from the
Catalan children in relationship to their Catalan identity and from the
non-Catalan Spanish (Andalusian) children in relationship to their Spanish
identity. In this case, the Spanish national group is numerically much larger
than the Catalan national group. However, Catalonia has achieved a high
level of political autonomy from the rest of Spain, has control of its own
educational, health, and social services, and uses Catalan as the official lan-
guage of education. Thus, the Spanish context is very different from the
British context in terms of the political influence and power exercised by
the Catalan people, and this needs to be borne in mind when interpreting the
findings from the Spanish context. The results of the analyses are shown in
the lower half of Table 6.8 (first two columns). It can be seen that a very
different picture emerges from the Catalan and Spanish children compared to
the Scottish and English children. And it is noteworthy that there is no real
evidence (except, counterintuitively, from the 6-year-olds) to support the
notion that the members of a minority group will identify with their national
group more strongly than the members of a majority group.

In-group homogeneity may be higher in members of minority

groups than in members of majority groups
Finally, the data from Barrett et al. (1997) were also examined to explore
whether the members of minority national groups perceive their national in-
group as being more homogeneous than members of majority national
groups. The expectation is that, within the British context, Scottish people, as
the minority and politically and economically subordinate group, should per-
ceive their national in-group as being more homogeneous and less variable
than should English people.
Lyons et al. (1997) used the data from the Scottish and English children in
order to test this idea. The findings are shown in the final column of Table 6.8
(upper half). It can be seen that, in the two younger age groups, there was no
difference in the perceived variabilities of the two in-groups. However, in the
12- and 15-year-olds, the predicted effect did occur: In both cases, the Scottish
children perceived their in-group to be more homogeneous than the English
children. Once again, the fact that this trend was exhibited only by the
older children, and not by the younger children, does make sense in terms of
182 Development of national identity
it being only the older children who are likely to have acquired sufficient
knowledge concerning the political, economic, and minority status of Scot-
tish people within Britain for this effect to occur. This finding underlines the
point that knowledge about relative group size and status are likely to mediate
the impact of social identity processes in children just as much as in adults (cf.
Bigler et al., 2001; Nesdale & Flesser, 2001).
However, the empirical findings are rather more complex than even this
might suggest. The comparable data from the Spanish context are shown in
the final column of Table 6.8 (lower half). A very different picture emerges
here. In this case, it is the younger members of the majority national group
who see their in-group as being more homogeneous; there is no significant
difference between the perceptions of the Catalan and Andalusian children at
15 years of age. Despite the necessary caveats concerning these data drawn
from the Spanish context, it is difficult to avoid the conclusion that the effects
of minority status do not occur consistently across children who are growing
up in different national contexts, and that different developmental patterns
may be exhibited depending on the specific national group to which the
child belongs.

SO DO SIT AND SCT PROVIDE USEFUL HEURISTIC

FRAMEWORKS FOR DEVELOPMENTAL RESEARCH?

Earlier in this chapter, it was noted that although SIT and SCT are not
developmental theories, it might nevertheless be possible to obtain some
insight into the development of national identity by examining whether
particular phenomena are exhibited by children at particular points in
development; the presence of these phenomena could then be interpreted as
evidence that the social identity processes postulated by SIT or SCT as being
responsible for the production of those phenomena have become operative in
these children. However, as we have seen, the findings are actually more
complex than this. Rather than there being evidence that particular phenom-
ena start to be exhibited by children at particular points in development,
pervasive differences have been found in the developmental patterns of
phenomena that are exhibited by children who are growing up in different
national contexts.
Thus, even the most basic phenomenon of in-group favouritism reveals
different developmental patterns in children who are growing up in Catalonia,
Scotland, and Italy (see Table 6.1). Similarly, while the strength of national
identification does indeed correlate with both positive and affective distinct-
iveness in Spanish children at all ages, it does not do so consistently in other
national groups that have been studied, and in some national groups, such a
correlation does not appear to be present at any age (see Tables 6.4 and 6.6).
Furthermore, differences (in strength of national identification, and in per-
ceptions of in-group homogeneity) between children who belong to minority
 Barrett, Lyons, del Valle 183
and majority national groups also seem to vary as a function of age in differ-
ent ways in different national contexts (see Table 6.8). In fact, the only finding
which seems to appear consistently in all national groups is affective prefer-
ence for the national in-group (see Table 6.2), which appears to be present
from 6 years of age onwards.
So do SIT and SCT provide useful heuristic frameworks for research into
identity development? If we are concerned with a close match between the
concrete predictions that many researchers have construed these theories as
making and the specificities of the data collected to date, and in using the
developmental patterns in these data to help delineate a general model of
identity development, the answer would appear to be no. However, if we are
concerned with the articulation of further hypotheses for empirical investiga-
tions in the future, the answer is, perhaps somewhat paradoxically, yes. That is
to say, SIT and SCT may still prove to be useful theoretical frameworks for
the articulation of further hypotheses concerning the development of
national identity, for the following reasons.
Turner (1999; Turner & Onorato, 1999) has recently argued that the core
social identity processes postulated by SIT and SCT (as summarised earlier in
this chapter) are likely to interact with a number of other factors, including
the individual’s motives, values, needs, goals, expectations, background
knowledge, theories and beliefs concerning the specific social groups
involved, and theories and beliefs about the prevailing intergroup relations
and the relative status and position of the in-group within the social system.
Furthermore, it seems plausible that all of these factors would be especially
important in the case of real-world social groups (as opposed to artificial,
experimenter-imposed, social groups). Thus, the extent to which particular
identity phenomena are exhibited may actually be a product of a much more
complex interaction between, for example, the strength of subjective identifi-
cation with the in-group, the individual’s beliefs about the nature of the
group boundaries (e.g., whether they are legitimate or illegitimate) and about
the status of the different groups within that system (e.g., whether the in-
group is high or low status, and whether this status is secure or insecure), and
the individual’s own personal motivations, values, needs, and expectancies
in relationship to these beliefs. In other words, Turner has argued that it may
actually be inappropriate to look for simple correlations between pairs of
variables.
Turner’s suggestion may help to explain why the identity phenomena that
are exhibited by children in the national identity domain vary so dramatically
according to the specific national context within which the children are grow-
ing up. For example, as has been noted already in this chapter, the political
realities of the intergroup relationships differ in Britain and in Spain.
Furthermore, the collective ideologies and shared beliefs concerning inter-
group relations (between Scotland and England within Britain, and between
Catalonia and Castilian Spain within Spain) also differ between the two
countries (e.g., language use – Catalan vs. Castilian – is one of the key issues
184 Development of national identity
within the Catalan context, but is far more marginal within the Scottish
context, where the political and economic subordination of Scotland to
England is probably the most significant issue). Therefore children who are
growing up in these different countries will be exposed to, and hence may
internalise, quite different beliefs about the prevailing intergroup relations
and about the relative status and position of their own in-group within the
national system. If, however, identity phenomena are a product of an inter-
action between, for example, the intensity of in-group identification and the
individual’s beliefs about the prevailing intergroup relations, then this might
explain the different patterns of development that are exhibited by the chil-
dren living in these two different national contexts.
If Turner is correct concerning the number of factors that are actually
responsible for eliciting particular identity phenomena, then there are two
implications for the future investigation of national identity development in
children. One of these implications is positive and the other is negative as far
as the utility of SIT and SCT as heuristic frameworks for developmental
research is concerned. The positive implication is that the research agenda
opens up dramatically. If identity phenomena really are an interactive
product of so many factors, then there are clearly many more variables
that need to be measured in developmental studies than have been
measured to date in any study. And it will be imperative for researchers
conducting such research to develop suitable methods of operationalising
and measuring all of these different variables in children, and to use a much
more complex multivariate approach to analysis, in order to explore properly
the implications of SIT and SCT.
However, the negative implication of Turner’s most recent formulations is
that they potentially render SIT and SCT impervious to empirical evalu-
ation. If negative findings that run counter to prediction are obtained, it is
possible that such findings can now always be explained post hoc by refer-
ence to some kind of cognitive or motivational factor. In other words,
precisely because so many loosely specified psychological factors (i.e.,
motives, values, needs, goals, expectations, background knowledge, theories,
and beliefs) have now been flagged as possible influences upon identity
phenomena, and because both SIT and SCT appear to have been aug-
mented by a number of auxiliary hypotheses that seem to form a protective
belt around the key ideas contained within these theories, it is possible that
these two theories have now been rendered invulnerable to empirical
refutation.
Certainly, the complexities of the existing data, alongside Turner’s most
recent statements, should caution us against drawing any oversimplistic
interpretations and conclusions concerning the role of social identity and
self-categorisation processes as far as the development of national identity is
concerned. It will be an issue for future researchers to decide, with the
advantage of hindsight, whether SIT and SCT have proved to be helpful and
productive heuristic theoretical frameworks for developmental research.
 Barrett, Lyons, del Valle 185
Acknowledgements
The study by Barrett, Lyons, Bennett, Vila, Giménez, Arcuri, and de Rosa
(1997), upon which this chapter has been largely based, was supported by a
grant received from the Commission of the European Communities DGXII,
Human Capital and Mobility (Networks) Programme (Grant No. CHRX-
CT94-0687), which was awarded to the Universities of Surrey, Dundee,
Girona, Málaga, Padova, and Roma “La Sapienza”. The following
colleagues contributed to the design of the research: Mark Bennett, Fabio
Sani, Ignasi Vila, Santi Perera, Almudena Giménez de la Peña, Luciano
Arcuri, Anna Emilia Berti, Annamaria Silvana de Rosa, and Anna Silvia
Bombi. We are extremely grateful to all of these individuals for their con-
tributions to this study. However, none of them should be blamed for the
theoretical speculations that have been put forward in this chapter.

REFERENCES

Aboud, F. (1988). Children and prejudice. Oxford: Blackwell.

Barrett, M. (1996). English children’s acquisition of a European identity. In G.
Breakwell & E. Lyons (Eds.), Changing European identities: Social psychological
analyses of social change (pp. 349–369). Oxford: Butterworth-Heinemann.
Barrett, M. (2001). The development of national identity: A conceptual analysis and
some data from Western European studies. In M. Barrett, T. Riazanova, & M.
Volovikova (Eds.), Development of national, ethnolinguistic and religious identities in
children and adolescents (pp. 16–58). Moscow: Institute of Psychology, Russian
Academy of Sciences (IPRAS).
Barrett, M. (in press). Children’s knowledge, beliefs and feelings about nations and
national groups. Hove, UK: Psychology Press.
Barrett, M., Day, J., & Morris, E. (1990, June). English children’s conceptions of
European people: a pilot study. Paper presented at the Conference on Perception
and Cognitive Processes, Trieste, Italy.
Barrett, M., & Farroni, T. (1996). English and Italian children’s knowledge of
European geography. British Journal of Developmental Psychology, 14, 257–273.
Barrett, M., Lyons, E., Bennett, M., Vila, I., Giménez, A., Arcuri, L., & de Rosa, A. S.
(1997). Children’s beliefs and feelings about their own and other national groups in
Europe. Final Report to the Commission of the European Communities, Director-
ate-General XII for Science, Research and Development, Human Capital and
Mobility (HCM) Programme, Research Network No. CHRX-CT94-0687.
Barrett, M., & Short, J. (1992). Images of European people in a group of
5–10-year-old English school children. British Journal of Developmental Psychology,
10, 339–363.
Barrett, M., Wilson, H., & Lyons, E. (1999). Self-categorization theory and the
development of national identity in English children. Poster presented at Biennial
Meeting of SRCD, Albuquerque, USA.
Bennett, M., Lyons, E., Sani, F., & Barrett, M. (1998). Children’s subjective
identification with the group and in-group favoritism. Developmental Psychology,
34, 902–909.
186 Development of national identity
Bernal, M. E., & Knight, G. P. (Eds.) (1993). Ethnic identity: Formation and transmis-
sion among Hispanics and other minorities. Albany, NY: State University of New
York Press.
Bigler, R. S., Brown, C. S., & Markell, M. (2001). When groups are not created equal:
Effects of group status on the formation of intergroup attitudes in children. Child
Development, 72, 1151–1162.
Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorization and the
formation of intergroup attitudes in children. Child Development, 68, 530–543.
Bourchier, A., Barrett, M., & Lyons, E. (2002). The predictors of children’s geo-
graphical knowledge of other countries. Journal of Environmental Psychology,
22, 79–94.
Branscombe, N. R., Ellemers, N., Spears, R., & Doosje, B. (1999). The context and
content of social identity threat. In N. Ellemers, R. Spears, & B. Doosje (Eds.),
Social identity (pp. 35–58). Oxford: Blackwell.
Branscombe, N. R., & Wann, D. L. (1994). Collective self-esteem consequences of
out-group derogation when a valued social identity is on trial. European Journal
of Social Psychology, 24, 641–657.
Brown, R., & Wootton-Millward, L. (1993). Perceptions of group homogeneity
during group formation and change. Social Cognition, 11, 126–149.
Eckes, T., & Trautner, H. M. (Eds.) (2000). The developmental social psychology of
gender. Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Ellemers, N., Doosje, B. J., Van Knippenberg, A., & Wilke, H. (1992). Status
protection in high status minorities. European Journal of Social Psychology, 22,
123–140.
Ellemers, N., Kortekaas, P., & Ouwerkerk, J. (1999). Self-categorization, commitment
to the group and social self-esteem as related but distinct aspects of social identity.
European Journal of Social Psychology, 29, 371–389.
Grant, P. R. (1992). Ethnocentrism between groups of unequal power in response to
perceived threat to valued resources and to social identity. Canadian Journal of
Behavioural Science, 24, 348–370.
Grant, P. R. (1993). Reactions to intergroup similarity: Examination of the
similarity-differentiation and similarity-attraction hypothesis. Canadian Journal of
Behavioural Science, 25, 28–44.
Haslam, S. A. (2001). Psychology in organizations: The social identity approach.
London: SAGE Publications.
Haslam, S. A., Oakes, P. J., Turner, J. C., & McGarty, C. (1995). Social categorization
and group homogeneity: Changes in the perceived applicability of stereotype con-
tent as a function of comparative context and trait favourableness. British Journal
of Social Psychology, 34, 139–160.
Haslam, S. A., Turner, J. C., Oakes, P. J., McGarty, C., & Hayes, B. K. (1992). Context-
dependent variation in social stereotyping 1: The effects of intergroup relations as
mediated by social change and frame of reference. European Journal of Social
Psychology, 22, 251–278.
Hinkle, S., & Brown, R. (1990). Intergroup comparisons and social identity: Some
links and lacunae. In D. Abrams & M. A. Hogg (Eds.), Social identity theory:
Constructive and critical advances. Hemel Hempstead, UK: Harvester Wheatsheaf.
Hopkins, N., & Cable, I. (2001). Group variability judgments: Investigating the
context-dependence of steretoypicality and dispersal judgments. British Journal of
Social Psychology, 40, 455–470.
 Barrett, Lyons, del Valle 187
Hopkins, N., Regan, M., & Abell, J. (1997). On the context-dependence of national
stereotypes: Some Scottish data. British Journal of Social Psychology, 36, 553–563.
Jahoda, G. (1962). Development of Scottish children’s ideas and attitudes about other
countries. Journal of Social Psychology, 58, 91–108.
Jaspers, J. M. F., Van de Geer, J. P., Tajfel, H., & Johnson, N. (1972). On the develop-
ment of national attitudes in children. European Journal of Social Psychology, 2,
347–369.
Johnson, N. (1973). Development of English children’s concept of Germany. Journal
of Social Psychology, 90, 259–267.
Johnson, N., Middleton, M., & Tajfel, H. (1970). The relationship between children’s
preferences for and knowledge about other nations. British Journal of Social and
Clinical Psychology, 9, 232–240.
Kelly, C. (1988). Intergroup differentiation in a political context. British Journal of
Social Psychology, 27, 321–327.
Lambert, W. E., & Klineberg, O. (1967). Children’s views of foreign peoples: a cross-
national study. New York: Appleton-Century-Crofts.
Lyons, E., Barrett, M., Bennett, M., Sani, F., Vila, I., Giménez de la Peña, A., Arcuri,
L., de Rosa, A. S., & Bombi, A. S. (1997, September). Intragroup homogeneity effects
in the development of national identity: A cross-national study. Paper presented
at the 8th European Conference on Developmental Psychology, Rennes, France.
Middleton, M., Tajfel, H., & Johnson, N. (1970). Cognitive and affective aspects of
children’s national attitudes. British Journal of Social and Clinical Psychology, 9,
122–134.
Mummendey, A., Klink, A., & Brown, R. (2001). Nationalism and patriotism:
National identification and out-group rejection. British Journal of Social Psych-
ology, 40, 159–172.
Nesdale, D., & Flesser, D. (2001). Social identity and the development of children’s
group attitudes. Child Development, 72, 506–517.
Oakes, P. J., Haslam, S. A., & Turner, J. C. (1994). Stereotyping and social reality.
Oxford: Blackwell.
Perreault, S., & Bourhis, R. Y. (1998). Social identification, interdependence and dis-
crimination. Group Processes and Intergroup Relations, 1, 49–66.
Phinney, J., & Rotheram, M. (Eds.) (1987). Children’s ethnic socialization: Pluralism
and development. Newbury Park, CA: SAGE Publications.
Piaget, J., & Weil, A. M. (1951). The development in children of the idea of the
homeland and of relations to other countries. International Social Science Journal,
3, 561–578.
Powlishta, K. K. (1995). Intergroup processes in childhood: Social categorization and
sex role development. Developmental Psychology, 31, 781–788.
Ruble, D. N., & Martin, C. L. (1998). Gender development. In W. Damon & N.
Eisenberg (Eds.), Handbook of child psychology, Vol. 3: Social, emotional and per-
sonality development (5th ed., pp. 933–1016). New York: John Wiley & Sons.
Sani, F., Bennett, M., & Joyner, L. (1999). Developmental aspects of contextual vari-
ability in the stereotype of the in-group: The case of Scottish children. Paper pre-
sented at the Annual Conference of the Social Psychology Section of the British
Psychological Society, Lancaster, UK.
Simon, B. (1992). The perception of in-group and out-group homogeneity:
Reintroducing the intergroup context. European Review of Social Psychology, 3,
1–30.
188 Development of national identity
Simon, B., & Brown, R. J. (1987). Perceived intragroup homogeneity in minority–
majority contexts. Journal of Personality and Social Psychology, 53, 703–711.
Simon, B., & Hamilton, D. L. (1994). Self-stereotyping and social context: The effects
of relative in-group size and in-group status. Journal of Personality and Social
Psychology, 66, 699–711.
Tajfel, H. (1978). Differentiation between social groups: Studies in the social psychology
of intergroup relations. London: Academic Press.
Tajfel, H., Jahoda, G., Nemeth, C., Campbell, J., & Johnson, N. (1970). The
development of children’s preference for their own country: A cross-national
study. International Journal of Psychology, 5, 245–253.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behaviour.
In S. Worchel & W. G. Austin (Eds.), Psychology of intergroup relations (2nd ed.,
pp. 7–24). Chicago, IL: Nelson-Hall.
Turner, J. C. (1999). Some current issues in research on social identity and self-
categorization theories. In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social iden-
tity (pp. 6–34). Oxford: Blackwell.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorization theory. Oxford: Blackwell.
Turner, J. C., & Onorato, R. S. (1999). Social identity, personality and the self-concept:
A self-categorization perspective. In T. R. Tyler, R. M. Kramer, & O. P. John (Eds.),
The psychology of the social self (pp. 11–46). Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Verkuyten, M. (2001). National identification and intergroup evaluation in Dutch
children. British Journal of Developmental Psychology, 19, 559–571.
Wilson, H. (1998). The development of national identity in 5 to 11 year old English
schoolchildren. Unpublished PhD Thesis, University of Surrey.
Wilson, H., Barrett, M., & Lyons, E. (1995, July). English children’s beliefs and feelings
about national groups in Europe: A pilot study. Paper presented at the Second
European Workshop on Children’s Beliefs and Feelings about their Own and Other
National Groups in Europe, University of Girona, Spain.
Yee, M. D., & Brown, R. J. (1992). Self evaluations and intergroup attitudes in
children aged three to nine. Child Development, 63, 619–629.
7 Ethnic identity and social
context
Maykel Verkuyten

This chapter is intended to make a case for the inclusion of social context
when studying ethnic identity among minority (and majority) group children.
Most writers agree that the social context plays a central role in ethnic
identity development. However, few researchers have actually examined
social context variables. Indeed, Coll et al. (1996) argue that existing research
and developmental models are insufficiently specific for the study of racial
and ethnic minority populations. An understanding of minority groups
would require explicit attention to negative social circumstances, such as
racism and discrimination, in relation to concrete environmental influences.
They argue, for example, that very little systematic research has been done to
examine how school variables promote or inhibit the well-being and social
competency of minority group children. Similarly, Rotheram and Phinney
(1987, p. 14) argue that:

The importance and meaning of ethnic identity varies with the specific
context and with changes in the social milieu and will be more salient in
some situations than in others. Children’s exposure to situations in which
they are aware of their ethnicity will vary depending on their status as
minority or majority group members, as well as on the degree of ethnic
homogeneity or heterogeneity in their daily activities. For example,
ethnicity is likely to be more salient for one white child in a class of 20
Black peers than for the same child in a predominantly white classroom.

Hence, the context is considered a central factor in understanding ethnic

identity, but there are relatively few studies on, for example, experiences with
discrimination and school characteristics. In this chapter I discuss our
empirical research that examines contextual issues among children in late
childhood (9 to 12 years old) living in the Netherlands1.
The notion of context is variously defined across many psychological

1 Much of this research has been conducted together with Barbara Kinket and with Jochem
Thijs.
190 Ethnic identity and social context
paradigms. For example, context is taken to refer to the particular task or
activity in which children are engaged, such as the comparative context in
eliciting self and group evaluations or the public or private expression of
these evaluations. Furthermore, the notion of context is used for historical
and cultural circumstances, immigration conditions, and actual social situ-
ations, such as in schools and neighbourhoods. This chapter focuses on
cultural and status differences between ethnic groups and actual local condi-
tions. It will be argued that an understanding of ethnic minority identity
requires explicit attention to cultural characteristics in addition to negative
social circumstances, such as discrimination, and in relation to concrete
environmental influences. Hence, both the “ethnic” and the “minority”
aspect of ethnic minorities are considered. The former aspect is typically not
examined by social psychologists, who focus on minority status and the
related issues of prejudice and discrimination. The ethnic aspect is stressed in
cross-cultural psychology and by those researchers who examine ethnicity
in terms of cultural differences. Both of these aspects are likely to play a role
in the development of ethnic identity, but they are typically not examined in
relation to each other.
The notion of context is not only variously defined but also addressed from
different theoretical perspectives. For example, in line with the symbolic
interactionist perspective, some theories focus on how socialisation experi-
ences with significant others and the wider sociocultural context influence the
way children come to view themselves ethnically (see Phinney & Rotheram,
1987). In this chapter, ideas of social psychological intergroup theories will be
drawn upon, particularly social identity theory (SIT; Tajfel & Turner, 1986)
and its elaboration self-categorisation theory (SCT; Turner, Hogg, Oakes,
Reicher, & Wetherell, 1987; Turner, Oakes, Haslam, & McGarty, 1994). I will
try to show that these theories offer useful and important frameworks for
studying ethnic identity among children, but that they also have limitations.
The first section gives a short historical description of psychological per-
spectives on the development of ethnic identity. The second section discusses
the (for this chapter) most relevant ideas of SIT and SCT. The third and
fourth sections present empirical results concerning perceived discrimination
and cultural orientations, and their relationships with ethnic identity. In the
fifth and sixth sections, the issue of the context dependency of ethnic identity
will be addressed. In the final section, a key aspect of SCT is examined,
namely the idea that children’s perceptions and evaluations are affected by
identity change or difference in self-categorisations.

FROM SOCIAL COGNITION TO SITUATION

Inquiry into the development of racial and ethnic identity has a well-
established research tradition in psychology. The early research by E. L.
Horowitz (1939), Clark and Clark (1947), and Goodman (1952) inspired
 Verkuyten 191
many studies. Empirically, a great deal of effort has been directed at
describing the ontogeny of children’s awareness of, identification with, and
preference for racial and ethnic categories (see Aboud, 1988; Fishbein, 1996,
for reviews).
Theoretically, different explanations have been offered in trying to account
for the appearance of racial and ethnic identity. Some major theories focus on
qualitative differences in racial and ethnic self-understanding that are associ-
ated with broad stages of cognitive development. An age-related progression
in the ability to perceive and interpret ethnic stimuli and inter-ethnic
behaviours is assumed (e.g., Aboud, 1987; Ramsey, 1987). The focus is on
processes and abilities, and the theoretical result is a sequence of steps or
stages in the development of aspects of ethnic identity2. The emphasis is very
much on the individual child and racial and ethnic identity is predominantly
examined as a psychological attribute. These cognitive theories have a strong
“individualistic” perspective in which there is little consistent simultaneous
examination of individual and contextual variables.
However, developments in the structural complexity of children’s thinking
say little about the particular beliefs and knowledge children acquire or
express. Racial and ethnic self-understandings are also determined by social
beliefs that are related to historical, cultural, and sociorelational contexts.
These social circumstances are important for examining and interpreting
possible differences between ethnic majority and minority group children
(Vaughan, 1987). Early studies conducted in America saw black children as
more or less passive recipients of the existing prejudice and stigmatisation to
which blacks as a group are subject. The core idea was that minority group
members come to internalise society’s negative view about their group and
therefore show the “mark of oppression” (Kardiner & Ovesey, 1951). Studies
found that black children were more ambivalent about their racial identity
than were whites, sometimes also showing preference for and identification
with the white out-group (see Brand, Ruiz, & Padilla, 1974; Milner, 1983, for
reviews). Although these studies have been criticised on methodological and
theoretical grounds (e.g., Banks, 1976; Brand et al., 1974), research in other
settings, such as Hong Kong (Morland, 1969), New Zealand (Vaughan,
1964), and Britain (Milner, 1973) has reported similar results.
These differences between minority and majority group children were typ-
ically interpreted in terms of the nature of inter-ethnic relations in society
and existing social structure. Social circumstances in the wider society were
also used to interpret the results of later studies conducted in America that

2 The cognitive development perspective has most often been used for generating theories about
ethnic identity development in children. Erikson’s work on identity development has also been
applied to ethnic identity but mainly to account for changes beyond childhood (e.g., Phinney,
1989; Weinreich, 1986). The same is true for the work of Cross (1991), who has proposed a
stage-model based on the African-American experience.
192 Ethnic identity and social context
showed a change in black children’s racial self-identification, from the former
ambivalence to a more consistent in-group identification and preference (e.g.,
Hraba & Grant, 1970; Mahan, 1976)3. This change was interpreted in terms
of the changing situation of blacks in American society, expressed in “black
pride” and civil rights movements (e.g., Fine & Bowers, 1984). However, most
of these studies failed to examine social context variables as such. Changes in
the wider society were used to interpret empirical results, but there remained
the question as to how actual experiences in concrete settings affect children’s
ethnic identity (Rosenberg, 1979).
Examining the role of context more systematically is in agreement with
some theories of human development, and also with the growing social psy-
chological interest in the importance of situational conditions. In develop-
mental psychology the ecological theory of Bronfenbrenner (1979), for
example, gives priority to the environment. Bronfenbrenner stresses the
role of interrelated and nested environmental structures, such as family,
school, and society, in affecting the content and structure of development.
In genetic social psychology (e.g., Emler & Ohana, 1993) the emphasis
is on the construction of knowledge as a social process, and on meanings
as social products; it is argued that theories of social cognition must go
beyond the individualistic doctrine of the cognitive-developmental
approach. And in social psychology, groups and group membership are seen
as deriving their identity from a particular context, and approaches such as
self-categorisation theory try to give a systematic account of the role of
situational conditions.

INTERGROUP THEORIES

Social identity theory (SIT) and self-categorisation theory (SCT) offer many
possibilities for examining children’s social self-definitions and understand-
ing of existing social relationships, in addition to psychological processes
such as categorisation and the need for positive self-esteem. However,
developmental studies using these theories have typically focused on inter-
group relations rather than social identity as such, and have not examined the
actual normative context, nor the role of status and cultural differences
(e.g., Bennett, Sani, Lyons, & Barrett 1998; Bigler, Jones, & Lobliner, 1997;
Nesdale, 2000; Powlishta, Serbin, Doyle, & White, 1994; Rutland, 1999;
Yee & Brown, 1992).
Because SIT and SCT are presented extensively in Chapter 1, I will focus
only on those aspects that are most relevant to the present discussion of
ethnic identity. The starting point for the theories is the distinction between

3 When comparing earlier (Milner, 1973) and later studies (Davey & Mullin, 1980), a similar
trend can be found in Britain.
 Verkuyten 193
personal and social identity. Personal identity refers to those idiosyncratic
characteristics and qualities that define the individual in relation to other
people, particularly in-group members. Social identity refers to social
category memberships. Both personal and social identity are considered
equally valid and authentic conceptions of the self and refer to different
levels of self-definition. People are both individuals and social group
members, and self-definition at the personal level and the group level is,
psychologically speaking, equally real. The relevance of the distinction
between personal and social identity lies in its differential consequences for
perception, evaluation, and behaviour (see Brewer & Miller, 1996; Oakes,
Haslam, & Turner, 1994, for reviews).
In conceptualising social identity, SIT and SCT have somewhat different
emphases. SIT stresses that social identity is connected with and derives from
the membership of emotionally significant social categories or groups (Tajfel,
1978), whereas in SCT the emphasis is on self-definitions. Hence, SIT stresses
more the motivational and affective aspects of social identities, whereas SCT
emphasises the situational and cognitive aspects. SCT is more concerned with
the contextually sensitive ways in which self-categorisations become salient
and less with group identification or the degree to which group membership
is psychologically central and valued. Hence, the concept of ethnic self-
categorisation emphasises the significance and consequences of a child’s
ethnicity within a particular context, whereas the concept of ethnic identifica-
tion emphasises individual differences in the degree to which ethnicity
develops in a psychologically central and valued group membership.
SCT does not deny the existence of more stable individual differences in
the tendency to define oneself in group terms (Turner, 1999). Some people
are more inclined than others to see themselves as a group member and some
value their group membership more than others (e.g., R. Brown & Williams,
1984). From a self-categorisation perspective, measures of identification are
useful for assessing the centrality and emotional value attached to a group
membership. Group identification is an important factor affecting a person’s
readiness to use a social category for self-description. Identification reflects
one of the psychological resources – together with individual motives, needs,
and goals – that are used to make sense of oneself and others in a particular
context. The momentary salience of particular self-categorisations is argued
to be determined by the individual’s readiness in interaction with the “fit”
between situational cues in the real world and the normative expectations
about group differences.
However, SCT’s emphasis on context and variation implies that the theory
has relatively little concern with explaining individual differences in group
identification or the affective commitments to, in our case, the ethnic group.
SIT and ideas developed in cross-cultural psychology are useful here.
Tajfel (1978) addressed the social psychology of ethnic minorities. He saw
the unfavourable social position of a group as the defining principle of
minority groups, and as the central issue for understanding minority identity.
194 Ethnic identity and social context
That is, the “minority” aspect of ethnic minorities is considered central.
Tajfel distinguished between numerical and psychological minorities, and
defined the latter as a group that feels bound together by common traits that
are held in low regard. He focused on the status differences between the
majority and the minority group and addressed the question of the psycho-
logical effects of minority membership with respect to the threat to social
identity that a minority position implies. He described how, depending on the
perceived legitimacy and stability of the social system, individuals can accept
or reject a negative social identity, and how minority groups may alter the
valuation of their group through creativity or social competition (Tajfel &
Turner, 1986).
Following this conceptualisation, social and developmental psychologists
have investigated children’s ethnic minority identity as an example of the
more general effect of status differences between groups and its concomitant
disadvantages, victimisation, and stigmatisation (e.g., Corenblum & Annis,
1993; Nesdale, 2001). Being a minority group member is seen as a threat to a
positive social identity. Children can respond to this threat differently: One
response is cross-ethnic identification, whereby minority group children dis-
tance themselves from their ethnic group and by doing so improve their self-
feelings. Another response is accentuating positively valued differences
between the in-group and out-groups, and strong and more lasting ethnic
group identification (Branscombe, Schmitt, & Harvey, 1999; Cross, 1991).
Authors using SIT typically see ethnic awareness, identification, and group
evaluations as reactions or responses to status differences and the predica-
ments of negative stereotypes, discrimination, and forms of racism. However,
there are also limits to using an exclusively social position perspective. Ethnic
identity is not composed simply of a minority status; treating it as such
greatly limits the ability to examine and understand the richness of the mean-
ings and experiences associated with this identity (Sellers, Smith, Shelton,
Rowley, & Chavous, 1998). In focusing on the “minority” aspect, ethnic
minority groups are treated as any low-status or powerless group to which the
same social-psychological processes are applied. As a result, the “ethnic”
aspect is ignored and there is a failure to theorise ethnicity.
In the social sciences, the notion of ethnicity is conceptualised in many
different ways. However, almost all link ethnicity to history and culture,
although in different ways (see Cornell & Hartman, 1998). Many social scien-
tists follow Max Weber (1968) in emphasising that ethnicity is primarily a
sense of belonging to particular (assumed) ancestry and origin (e.g., Cornell
& Hartman, 1998; DeVos, 1995; Roosens, 1994). An ethnic group is thought
to exist whenever the belief in common descent is used to bind people
together to some degree. This sense of origin is often accomplished by
defining ethnicity in terms of metaphors of kinship: Ethnicity is family writ
large (D. L. Horowitz, 1985). In addition to ancestry, ethnicity is most often
thought of as culture that is transmitted across generations.
Culture offers an important framework for interpreting the world.
 Verkuyten 195
Members of different cultures sample and interpret the environment
differently. Many studies emphasise various cultural characteristics that
shape perception and behaviour (e.g., Hofstede, 1980; Markus, Kitayama, &
Heiman, 1999; Smith & Bond, 1993). For example, the most widely examined
distinction in cross-cultural studies currently is between collectivism and
individualism, which is studied not only at the cultural level but also at the
individual level (Singelis, 1994; Triandis, Leung, Villareal, & Clack, 1985).
Collectivism can be thought of as a worldview or interpretative frame that
functions to focus attention and thinking. A collectivist worldview promotes
the perception of self and others in terms of ascribed group membership,
such as family and ethnicity: The group is the centre of the psychological field
(Oyserman, 1993; Triandis, 1994). Enculturation in a more collectivist culture
implies a relatively high group consciousness and a strong predisposition to
identify with one’s own ethnic group.
In our research we have focused on individual and ethnic group differences
in perceived discrimination and in collectivism as a cultural value orientation.
Furthermore, various studies on ethnic identity (e.g., Garza & Herringer,
1987; Phinney, 1990; Sellers et al., 1998), and social identity in general (e.g.,
Ellemers, Kortekaas, & Ouwerkerk, 1999; Jackson & Smith, 1999; Smith,
Murphy, & Coats, 1999), have shown that different aspects of identity can be
distinguished. In SIT, social identity is linked to the value and evaluative
significance of group membership. This theory assumes that a group member
is motivated by a need for positive self-esteem as a group member. In this
chapter the focus is on this aspect of social identity for which the term ethnic
self-esteem will be used (see Ellemers et al., 1999). Finally, in our research,
characteristics of school settings have been used for examining the role of
local situational circumstances. In particular, the ethnic composition of schools
and normative issues related to multicultural education have been examined.

PERCEIVED DISCRIMINATION4

In social psychology, increasing attention is being paid to the phenomenology

of being a target of prejudice and discrimination (see Crocker, Major, &
Steele, 1998; Swim & Stangor, 1998). There is a growing interest in the
minority member’s or insider’s perspective (Oyserman & Swim, 2001).

4 Experiences with discrimination are discussed here as more enduring or pervasive features that
influence ethnic identification. However, discrimination can also be more incidental and situ-
ational (L. M. Brown, 1998; Verkuyten & Thijs, 2001) and therefore may affect contextual
self-categorisations. Furthermore, the direction of the relationships between discrimination
and identification is not fixed: Ethnic identification may also affect the perception of dis-
crimination (Phinney, Madden, & Santos, 1998). For example, if a child perceives discrimin-
ation, he or she may suffer from a more negative sense of ethnic identity, which, in turn, may
lead to increased sensitivity to others and to behaviour that elicits discriminatory responses
from others.
196 Ethnic identity and social context
However, these issues are typically examined with university student and
adult samples, and not with children.
In a number of studies among older children we have investigated the issue
of discrimination and its relationship with ethnic self-esteem. We started our
investigation by examining children’s own understandings of discrimination
(Verkuyten, Kinket, & Van der Wielen, 1997). Using open-ended instruments
and short stories, we found shared beliefs and ideas about when a specific act
is considered discriminatory among both Dutch and ethnic minority chil-
dren. According to the children, the prototypical example of discrimination
was a situation in which a Dutch child called a minority child names. To a
lesser degree, an unequal division of valued objects among contemporaries
and social exclusion from play by peers were also seen as discrimination. In
short, children’s understanding about discrimination was predominantly
in terms of peer relations and not on the level of the wider society. This
understanding may be important for ethnic self-esteem. Studies by Harter
(1999) have shown that (dis)approval from peers, such as classmates, is far
more predictive of self-evaluations than is (dis)approval from one’s close
friends. Furthermore, in general, social exclusion is more strongly related to
self-evaluation than is social inclusion (Leary & Baumeister, 2000).
A second step in our research was to investigate the perception of dis-
crimination (in terms of ethnic name-calling and social exclusion by peers) by
ethnic minority and Dutch children. In different studies we found a clear
difference in perceived discrimination between both groups of children
(Verkuyten, 2001; Verkuyten & Thijs, 2002). Turkish, Moroccan, and
Surinamese children felt more discriminated against than did Dutch children.
In addition, Turkish children reported higher discrimination than the other
ethnic minority groups. This is in agreement with studies showing that the
Turks are the group that is least liked by Dutch adolescents and adults
(Hagendoorn, 1995) and by older children (Verkuyten & Kinket, 2000).
These results show that not all ethnic minority groups face the same level of
derogation and social exclusion. Ethnic minority groups are distinguishable
from others in terms of their relationship to the majority group.
In different studies we further found that minority and majority group
children perceive a higher level of discrimination directed at their group as a
whole than at themselves as individual members of that group (Verkuyten,
2002b). Taylor, Wright, Moghaddam, and Lalonde (1990) have labelled this
phenomenon the personal/group discrimination discrepancy (PGD). This
discrepancy is a robust finding among an array of (disadvantaged) groups
and using different wordings of questions (see Taylor, Wright, & Porter,
1993). Several explanations have been offered for this phenomenon, such as
the denial of personal discrimination, the exaggeration of discrimination
directed at the group as a whole, and basic features of information process-
ing. Whatever the explanation, the phenomenon seems real and our results
show that it is not restricted to students and adults. Children in late childhood
also make a clear distinction between the personal and group level. This has
 Verkuyten 197
also been found by Rosenberg (1979), who established that children may
agree with ethnic stereotypes about their group without believing that these
characterise them personally. Thus, although we found minority group
children to perceive far more discrimination both directed at their group and
at themselves as members of that group, both minority and majority group
children showed the discrepancy between personal and group discrimination.
Most authors argue for a close relationship between ethnic identity and the
existence of prejudice and discrimination. Discrimination is considered one
of the major factors affecting ethnic minority identity (e.g., Cross, 1991;
Keefe, 1992). In our studies we have found a clear negative relationship
between perceived discrimination (name-calling and social exclusion) and
ethnic self-esteem, particularly among Turkish children (Verkuyten & Thijs,
2000, 2002b). When children experience more ethnic discrimination they
evaluate their ethnic identity less positively. Furthermore, personal dis-
crimination was found to affect ethnic identification negatively (Verkuyten,
2002b). In addition, perceived group discrimination affected ethnic self-
esteem negatively, but only among children who considered their ethnic
identity an important part of the self.
Thus, ethnic name-calling and exclusion from play because of one’s eth-
nicity are more common experiences for ethnic minority group children and
have a negative impact on their ethnic self-evaluation. This does not imply,
however, that for these children ethnic identity is psychologically less central
and more negative compared to the majority group. On the contrary, dis-
crimination can be one of the reasons why ethnicity is a very central identity
(Keefe, 1992). Numerous authors have argued that ethnic identity is crucial or
at least central to the self-concept and psychological functioning of ethnic
minorities (see Cross, 1991; Liebkind, 1992; Phinney, 1990). We have found in
several studies that ethnic identity is more important for minority than for
Dutch children, and that discrimination has a modest but significant positive
correlation (around .20) with the importance that children attach to their
ethnic identity (Verkuyten & Thijs, 2000).
In addition, different studies in the Netherlands have consistently found
that children of ethnic minority groups evaluate their group membership
more positively and feel more committed to their group than do majority
group children (see Verkuyten, 1999, for a review). One possible explanation
for these findings is in terms of SIT. According to SIT, children who strongly
identify with their group are motivated to evaluate their own group positively
in comparison to other groups. For these children, the group membership
has important implications for the self-concept, and as such for the striving
for a positive self. Minority status and discrimination are considered to be a
threat to a positive ethnic identity and that threat can be counteracted by
accentuating positive distinctiveness and stronger in-group identification.
Empirical results, also in the Netherlands, indicate that children typically
demonstrate a positive bias towards their own ethnic group, and do not
necessarily evaluate out-groups negatively (Cameron, Alvarez, Ruble, &
198 Ethnic identity and social context
Fuligni, 2001; Verkuyten, 1999). Furthermore, among ethnic minority group
children, in-group identification and positive in-group evaluation are often
not accompanied by stronger intergroup differentiation. Ethnic identification
is typically related to in-group preference rather than to out-group dislike (see
Verkuyten, 1999).
There are some important assumptions to the SIT explanation of strong
group identification and positive in-group evaluation among ethnic minority
groups. One assumption is that ethnic minority children define and evaluate
their identity in relation to the dominant majority group (Vaughan, 1987). In
SIT, the emphasis is often on dichotomous and competitive intergroup com-
parisons whereby the majority group is implicitly assumed to be the only
really significant other (Spielman, 2000). However, minority group children
may value their ethnic identity because of supportive relationships with other
in-group members and because they prefer similarities over differences
(Brewer, 1999; Cameron et al., 2001; Fishbein, 1996). In addition, most ethnic
minority groups have their own rich history, culture, and traditions. These are
important sources for developing pride and satisfaction in one’s ethnic back-
ground and a positive ethnic identity (Hutnik, 1991). Ethnic minority groups
are “ethnic” from the inside and have their own social networks and sources
for a positive ethnic identity. Furthermore, ethnic minority group parents are
often in a position of striving to sustain cultural values and traditions that
are different from the values of the society in which they live. For example,
Turkish parents in the Netherlands are, in general, very concerned with
transmitting their traditions, history, and cultural values to their children
(e.g., De Vries, 1987; Nijsten, 1998). From birth on, the child’s intimate con-
tact is with members of his or her group. So, for example, Turkish children
know from early experience that stereotypes such as Turkish people being
unfriendly, dishonest, and lazy are not true. They learn to value Turkish
people and Turkish culture from early on and develop an intimate knowledge
of the way Turkish people are.
In a study among Turkish and Dutch participants, we found among the
former, but not among the latter group, a positive correlation between child-
ren’s ethnic self-esteem and parental intergroup attitudes (Verkuyten, 2002a).
Furthermore, ethnic minority group children who have poor cultural know-
ledge tend to reject their ethnic identity (Bagley & Coard, 1975). In addition,
similar to ethnic minority group differences in social status positions and
experiences with discrimination, there can also be cultural differences between
minority groups that may account for differences in ethnic identification. For
example, Milner (1973) found stronger in-group identification among Asian
than West Indian children living in Britain. He argued that “the more
autonomous, self-reliant Asian communities and the traditional family-life of
their members promote a more positive identity in the Asian child” (p. 293).
 Verkuyten 199
CULTURAL ORIENTATIONS

From the position of genetic social psychology, cognition must be seen as being
embedded in historical, cultural, and sociorelational contexts. People who are
members of the same culture or group will share the contents of thought in
the form of shared knowledge and beliefs (Emler, 1987; Moscovici, 1984).
Cognitions are not purely individual constructions, but are greatly influenced
by the kinds of beliefs available in the child’s surrounding environment.
Existing studies on self-identification assume that individual children differ
in their psychological group membership: Children identify more or less
strongly with their ethnic group. However, this assumption may not be
equally adequate for minority groups with a more collectivistic culture in
which the interdependence of in-group members is emphasised. In these cul-
tures group membership may be far less negotiable and “the question of why
individuals become committed to groups and identify with them is unlikely to
be asked” (Markus et al., 1999, p. 900). Children may become attached to a
group for interpersonal reasons, and being committed to the group can be
more a self-evident matter and an issue of duty and moral obligation. This is
particularly likely to happen with ethnic groups that are defined in terms of
common ancestry and family-origin metaphors (Roosens, 1994). However, in
studies among ethnic minority children, individual differences in ethnic
identification are nonetheless found and these may be related to individual
differences in cultural value orientations.
In different studies we have examined the issue of collectivism and its
relationship with ethnic self-esteem. Cultural tendencies towards collectivism
have different aspects including subordination of own goals to the goals of
others, concern for the in-group, interdependence and sociability, and family
connectedness. Collectivist cultures emphasise the interconnected nature of
the self, group solidarity and sharing, stable and predetermined relationships,
and family harmony and dependency (e.g., Triandis, 1994). Collectivism can
be studied both at the cultural and individual level. To distinguish clearly
between these levels, Triandis et al. (1985) refer to collectivism at the personal
level as allocentrism. In our studies we have predominantly focused on family
allocentrism, which is considered a central aspect of collectivism. Family
allocentrism is probably also one of the most meaningful dimensions for
children and its importance has been highlighted in cross-cultural work
(Kaǧitçibaşi, 1990). Furthermore, in the context of acculturation processes,
the family plays an important role and family allocentrism has been found to
have significant implications for immigrant experiences and ethnic identity
(Lay et al., 1998; Phinney & Rosenthal, 1992).
Ethnic minority groups in the Netherlands, such as Turks and Moroccans,
are originally from more collectivist cultures (Hofstede, 1980). Hence, as a
group Turkish and Moroccan children can be expected to endorse collectivist
values more strongly than the Dutch. In several studies we have indeed found
this to be the case (Verkuyten, 2001), and this is also found among young
200 Ethnic identity and social context
adolescents (e.g., Huiberts, Vollebergh, & Meeus, 1999; Verkuyten & Masson,
1996).
This group difference in collectivism may (partly) explain the finding that
compared to the majority group, minority group children perceive more
ethnic discrimination. Collectivism implies attentiveness to group members.
The group operates as a cognitive schema that influences perception with a
relatively high awareness and attention to information related to one’s group,
such as a heightened sensitivity to ethnic-relevant information. In two studies
we found a significant correlation between family allocentrism and perceived
discrimination (Verkuyten, 2002b). Independent of ethnic identification,
higher allocentrism was associated with more perceived personal and group
discrimination. This relationship was found for different ethnic groups,
including the Dutch. It suggests that for children high on allocentrism, the in-
group is part of the self and the group functions as a cognitive schema that
influences perception. Compared to children low on allocentrism, attention is
more focused on group-related information involving a greater attentiveness
and sensitivity to one’s ethnic group and ethnic group membership.
Using open-ended instruments, different studies among older samples have
examined group differences in collectivism in relation to self-descriptions.
These studies have found that collectivist groups give self-descriptions that
include more social identities, including ethnic identity (e.g., Cousins, 1989;
Rhee, Uleman, Lee, & Roman, 1995). However, there are only a few studies
among children (e.g., Rotenberg & Cranwell, 1989): An example is Van
Den Heuvel, Tellegen, and Koomen (1992), who found that Turkish and
Moroccan children referred more to social group memberships compared to
Dutch children, who gave more self-descriptions related to psychological
characteristics. Another example is our own research, in which we found that
Turkish children gave more ethnic self-descriptions than Dutch contemporar-
ies (Kinket & Verkuyten, 1997). Hence, cultural differences between ethnic
groups seem to be related to the type of ethnic self-descriptions. This is also
true at the level of the individual child. Controlling for discrimination, we
found among Turkish and Dutch children a positive association between
family allocentrism and ethnic self-descriptions (Verkuyten, 1999).
Ethnic group differences in collectivism may further explain (in part) the
differences found for ethnic self-esteem. The fact that, in general, ethnic
minority group children evaluate their ethnic identity more positively may be
due to their stronger collectivist worldview. This relationship can also be
examined at the individual level. Because allocentric individuals are more
likely to define themselves in relation to their in-group, evaluations of in-
group membership is particularly important for their well-being. In different
studies among ethnic minority and Dutch children we found a positive cor-
relation (around .30) between family allocentrism and ethnic self-esteem
(Verkuyten, 2001, 2002b). Moreover, this association was found after control-
ling for perceived discrimination. Hence, higher allocentrism was related to a
more positive ethnic identity. Furthermore, both discrimination and
 Verkuyten 201
allocentrism had independent effects on ethnic self-esteem, showing the
importance of taking both kinds of factors into account simultaneously.

ETHNIC SELF-ESTEEM AND CONTEXT

Identification with a group is typically conceptualised as an individual differ-

ence variable representing some kind of stable self-structure or trait-like
aspect that is expressed independently of the social situation (e.g., Crocker &
Luhtanen, 1990). The idea is that particular group memberships will develop
a central and more or less fixed place in children’s self-understanding. Hence,
identification is, by definition, relatively independent from context. This does
not mean, however, that children will always feel the same about their ethnic
group membership across situations. It only means that ethnic identity is not
strongly governed by situational influences. Situational variability will exist,
but gradually the child will probably also develop a more stable affective and
evaluative attitude towards his or her ethnic background. This would mean
that the value and emotional significance that children attach to their ethnic
identity will not differ much between situations.
To my knowledge there are no studies that have examined the temporal
stability of ethnic self-esteem among children, nor whether children evaluate
their ethnic identity differently across relational contexts. In our research we
have examined the context dependency of ethnic self-esteem by studying
classroom contexts. This research was conducted within classrooms so that
the class context was salient. In three studies involving more than 200 class-
rooms, we investigated ethnic self-esteem using multilevel modelling. Multi-
level modelling allows the simultaneous analysis of individual and group level
variables without compromising the quality of the information at any level
(Kenny, Kashy, & Bolger, 1998)5. In our research the samples consisted of
school classes and children within these classes: The children are nested
within their classes. Hence, ethnic self-esteem is explained by children’s indi-
vidual characteristics and by properties of the classroom. A significant effect
found at the level of the classroom indicates that the similarity in ethnic
self-esteem between children in the same classroom is greater than between
children in different classes.
In three different studies we found a very similar result (Verkuyten, 2001;

5 Multilevel analysis is also necessary because there are statistical disadvantages to handling
contextual variables at the individual level. Assigning group level variables to individuals may
lead to spurious significant results because the standard errors, which are based on the higher
number of disaggregated cases, are too small. Hence, originally small differences between
contexts will become significant because of an increased number of observations. Further-
more, groups, and in particular classes in school, are hardly ever formed randomly and chil-
dren that belong to the same class will share many experiences. Therefore, the assumption of
independence of observations is often violated (Kenny & Judd, 1984).
202 Ethnic identity and social context
Verkuyten & Thijs, 2003). In all three, the multilevel analysis indicated that
the class-level variance was significant for ethnic self-esteem. In some classes
children evaluated their ethnic identity more positively than in others. How-
ever, in all three, the percentage of the variance explained by the grouping
structure was low (between 6.9% and 8.2%). Hence, the within-class variance
(between 93.1% and 91.8%) was much larger than the between-class variance.
Therefore individual factors clearly explained more variance in ethnic self-
esteem than did classroom features6. Thus, although ethnic self-esteem was
affected by the school setting, it predominantly depended on individual char-
acteristics. This is not to say that ethnic self-esteem is some kind of fixed or
stable personality characteristic that is chronically salient across situations
and expressed independently of social circumstances. It merely suggests that
measures of ethnic self-esteem reflect individual differences in the affective
and evaluative involvement in one’s ethnic group and thereby in the child’s
readiness to self-categorise in ethnic terms.

ETHNIC SELF-CATEGORISATION

For SCT, the context dependency of self-categorisation is central. The theory

is concerned with explaining the extent to which, in our case, one’s ethnicity is
psychologically salient at a particular moment and in a particular situation.
Following this idea, it can be expected that children’s ethnic self-definitions
will strongly depend – and depend more strongly than ethnic self-esteem – on
the actual social situation.
In a study among Turkish and Dutch children we investigated this idea in
relation to the classroom context (Kinket & Verkuyten, 1997). Children were
asked to give 10 self-descriptions. Answers referring explicitly to ethnic group
membership were coded. Most of the children (65%) did not refer to their
ethnicity when asked to describe themselves. However, the multilevel analysis
indicated that the between-class variance was significant. As much as 44% of
the variance in ethnic self-descriptions was explained by the grouping struc-
ture. Hence, independently of individual differences, children were far more
likely in some classes than in others to refer to their ethnicity in their spon-
taneous self-descriptions. The salience of ethnicity was highly sensitive to the
context of the classroom situation. Thus, compared to ethnic self-esteem, the
self-definition depended more strongly on who was present and on the way
the context was defined. One classroom characteristic that affected ethnic

6 Of the different classroom characteristics, ethnic composition, as well as educational issues,

had an effect. For example, Turkish children evaluated their ethnic group membership more
positively when the percentage of Turkish classmates was higher. Ethnic self-esteem was also
more positive in classes where the importance of multicultural issues was stressed and teachers
reacted to incidences of ethnic harassment with sanctioning of the perpetrator. A full
discussion of these and other effects can be found in the articles.
 Verkuyten 203
self-description was the percentage of in-group pupils. The Turkish children
were more likely to use ethnicity in their self-descriptions when the percent-
age of Turkish classmates was higher. The trend was reversed for the Dutch
children: They were more likely to describe themselves in ethnic terms when
the number of Dutch pupils was lower.

According to SCT, self-categorisation involves depersonalisation of self-

perception leading to self-stereotyping or the perception of increased similar-
ity between the self and typical in-group members. As argued earlier, ethnic
self-esteem is a main factor determining self-categorisation. Thus, the extent
to which children value their group membership and commit themselves to
their group can be expected to be related to the perception of similarity
and prototypicality of the self in relation to the group. Children who identify
with or commit themselves strongly to their ethnic group may have a greater
tendency to self-stereotype.
In a study among Dutch and Turkish children we examined this expect-
ation (Verkuyten, 1999). Our focus was not on self-stereotyping in terms of
specific group traits and attributes but on the perception of general proto-
typicality. Self-stereotyping was assessed in terms of one’s general similarity
to other children of one’s ethnic group and the extent to which children
consider themselves a typical group member. Principal components analysis
showed that ethnic self-esteem is empirically distinguishable from self-
stereotyping. Further, Turkish children exhibited a significantly higher level
of self-stereotyping than the Dutch. However, among both groups there was
a similar significant association between ethnic self-esteem and self-
stereotyping (r = .38, p < .001). As expected, higher ethnic self-esteem was
related to stronger ethnic self-stereotyping. Hence, children who identify
strongly with their ethnic group had a greater tendency to self-stereotype.
In summary, compared to ethnic self-esteem, ethnic self-definition is
strongly dependent on the social context. The former seems to reflect the
centrality and affective meaning of one’s ethnic group membership, which
determines children’s readiness to define themselves in ethnic terms and to
see themselves as typical group members. However, whether the child in a
particular situation actually self-categorises in ethnic terms depended not
only on ethnic identification. Various personal needs, motives and goals are
important in interaction with the normative expectations about group differ-
ences and the distribution of stimuli in the social context, such as the number
of classmates from one’s own ethnic group.

ETHNIC SELF-CATEGORISATION AND BEHAVIOUR

Although differences in ethnic identification are of central interest to the

social identity perspective, self-categorisation is considered to constitute
the psychological basis for behaviour. According to SCT, social categories
204 Ethnic identity and social context
influence behaviour when individuals define themselves in terms of those
categories. It is argued that when a given identity is salient, children will think
and act in terms of those beliefs and standards that define the salient identity.
Hence, when identities change, beliefs and standards change and, therefore,
perceptions and judgments will also change. This section of the chapter
discusses two experimental studies that illustrate this process. In the first,
the distinction between personal and ethnic identity was used; the second
concerned social identities of bicultural children.

The first study set out to examine peer victimisation in relation to the
distinction between personal and social identity (see Verkuyten & Thijs,
2001). Studies on peer victimisation typically do not consider ethnic dis-
crimination (see Hawker & Boulton, 2000), whereas studies on peer group
discrimination do not consider other forms of victimisation. However, in
order to understand the precise role of discrimination, other forms of victim-
isation should be considered. A distinction between personal and ethnic vic-
timisation may be particularly useful. Personal victimisation refers to those
situations where negative peer experiences are related to individual character-
istics, such as acting “strange” and stuttering. Ethnic group victimisation
occurs when children’s negative experiences are connected to their ethnic
group membership.
This distinction between personal and group victimisation is related to the
distinction between personal and social identity. Applying the personal–
social distinction to feelings of self-worth, we focused on personal self-esteem
and ethnic self-esteem. Differences in personal and ethnic self-esteem were
expected to predict self-reported peer victimisation. In order to examine this
prediction, we elicited self-reports on experiences with either personal or
ethnic victimisation. Following SIT and SCT, self-reported victimisation was
expected to be related to self-esteem at the same level of abstraction. Hence, it
was predicted that personal self-esteem and not ethnic self-esteem would be
negatively related to self-reported personal victimisation. On the other hand,
ethnic self-esteem, and not personal self-esteem, was expected to be nega-
tively related to group victimisation. These predictions were tested in an
experimental questionnaire study among 106 Turkish children. As expected,
it was found that personal self-esteem negatively predicted personal victim-
isation but not ethnic victimisation, and ethnic self-esteem tended to predict
reports of ethnic victimisation but not personal victimisation.
In addition, to examine the causal effects of both personal and ethnic
victimisation, we assessed momentary self-feelings directly after self-reported
peer victimisation. It was expected that peer victimisation would have a nega-
tive effect on momentary self-feelings, independently of the level of personal
and ethnic self-esteem. Furthermore, in order to examine the possible differ-
ent effects of personal and ethnic victimisation, a between-subjects design
was used with two conditions. For one group of Turkish children, experiences
with personal victimisation were elicited, whereas for the other group ethnic
 Verkuyten 205
victimisation was made salient. We explored whether personal and ethnic
victimisation differ in their negative impact on momentary self-feelings. In
contrast to personal victimisation, ethnic victimisation is more specific for
ethnic minority children than for children of the majority group. Hence,
compared to personal victimisation, ethnic victimisation may be more
strongly related to negative self-feelings. The results showed that peer victim-
isation had a negative causal effect on momentary self-feelings independent
of the level of personal and ethnic self-esteem. In addition, peer victimisation
based on ethnic group membership had a somewhat stronger negative effect
on self-feelings than victimisation based on personal characteristics.

The second example is a study in which we examined the effect of social

identity salience on bicultural children’s self-evaluations. Ethnic minority
group children not only belong to their own ethnic group but are often also
citizens of the country where they were born, live, and where their future is.
These children have to deal with the expectations and demands of their own
ethnic group as well as those of the majority group. Beliefs and standards that
are valued and learnt within the ethnic in-group may differ from beliefs and
standards that are stressed in society and its institutions such as schools.
Children have to learn to adapt to this bicultural situation. Different models
have been proposed for describing and understanding these adaptation pro-
cesses, such as alternation and fusion models (see LaFromboise, Coleman, &
Gerton, 1993). However, little is known about why children shift between
these frameworks and what the consequences of these shifts are. One possibil-
ity is that this switching occurs in response to contextual cues that activate
different social identities.
SCT argues that when a given identity is salient, beliefs and standards that
define that salient identity will govern people’s thinking and acting. Hence,
changes in identities imply changes in beliefs and standards leading to
changes in perceptions and judgments. As argued earlier, ethnic differences
are predominantly understood in terms of culture. Many studies have found
ethnic group differences in cultural value orientations, such as collectivism
and individualism (e.g., Gaines et al., 1997; Rotenberg & Cranwell, 1989).
These cultural orientations influence perceptions and judgments, including
self-evaluations (see Markus et al., 1999). For example, in self-description
tasks, people in individualistic cultures are more likely than those in collectiv-
ist cultures to make self-enhancing statements. However, a self-evaluation
motive is not necessarily absent in collectivist cultures. Members of these
cultures are more likely to evaluate their social identities favourably, whereas
in individualist cultures, personal identities are more likely to be evaluated
positively (Hetts, Sakuma, & Pelham, 1999; Pelham & Hetts, 1999). Fur-
thermore, Hetts and colleagues have argued, and found among ethnic minor-
ity groups, that these tendencies are most evident for more indirect or implicit
measures of self-evaluation. These measures are thought to reflect people’s
nonconscious feelings about the self that are related to the normative beliefs
206 Ethnic identity and social context
and values learnt in early socialisation. Explicit self-evaluations are more
transparent and are probably more dependent on the current social and
cultural context (Greenwald & Banaji, 1995). Hence at the more conscious
level, ethnic minorities will probably endorse the kind of favourable
conceptions of the self that are promoted in individualistic cultures, such as
self-enhancement.
Following SCT and these findings on cultural differences, it can be
expected that bicultural children will evaluate themselves and their group
differently depending on whether their ethnic minority identity or Dutch
identity is salient. In particular, when ethnic identity is salient a more positive
implicit evaluation of social identity can be expected, whereas personal
identity will be more positive when Dutch identity is salient. For explicit
questions on self-enhancement few or no differences can be expected.
To test these predictions we conducted a study among 74 bicultural Greek-
Dutch children7 (Verkuyten & Pouliasi, 2002). Greek culture has been found
to be more collectivist than the Dutch one (Hofstede, 1980). For priming
ethnic and Dutch identity we used iconic cultural symbols and language.
That is, similar to the work of Hong, Morris, Chiu, and Benet-Martinez
(2000), we randomly presented Greek-Dutch children with either Dutch
icons (e.g., national flag, windmills) or Greek icons (e.g., national flag,
Acropolis) in the Dutch and Greek language respectively. The combination
of icons and language was considered an effective means of activating the
two different identities. For more implicit measures of personal and ethnic
self-evaluation we asked the children about their feelings about the words “I”
and “We” (Hetts et al., 1999). The results were in agreement with the expect-
ations. Compared to the Dutch identity condition, in the Greek condition,
the children had a significantly more positive response to “We” and a more
negative response to “I”. Hence, depending on the salient social identity, the
more implicit evaluation of the personal and social self differed. In contrast,
for the explicit measure of self-enhancement, no significant difference
between the two conditions was found.8

7 Dutch children and Greek children in Greece also participated in this study, for two reasons.
First, finding a difference in self-evaluations between the Greek and the Dutch identity
priming conditions does not necessarily reflect cultural identity switching. One alternative
interpretation for a more positive social identity evaluation in the Greek condition is that the
minority position of the in-group in the Netherlands is made salient. As argued by SIT, in such
a situation people can respond by emphasising their social identity and accentuating positive
in-group distinctiveness. A cultural interpretation is more convincing when the result for
Greek identity activation among bicultural children is similar to that for Greek children in
Greece, and the result for Dutch identity activation is similar to that for Dutch children. This
was found to be the case. Second, by including a sample of monocultural Dutch children and
Greek children in Greece, it can be examined whether there are indeed significant cultural
differences in collectivism between the children of both societies. This was also the case.
8 These results can, of course, be interpreted in different ways. One is in terms of an explicit
structural model in which the personal and social self are seen as cognitively represented and
stored in two distinct “baskets” (Trafimow, Triandis, & Goto, 1991). Another explanation is
 Verkuyten 207
CONCLUSION

This chapter has tried to make a case for the inclusion of social context when
studying ethnic identity among children. Most writers agree that the social
context plays a central role, but few empirical studies have addressed this
issue. The present discussion has been on cultural and status differences
between ethnic groups and actual local conditions. An understanding of eth-
nic minority identity requires explicit attention to cultural characteristics and
negative social circumstances, such as discrimination, in relation to concrete
environmental influences.
Theoretically, the development of ethnic identity is typically studied from
a social-cognitive perspective. Although the models differ somewhat, they
all describe an age-related progression in the ability to perceive and inter-
pret ethnic differences (see Aboud, 1988; Brown, 1995; Phinney &
Rotheram, 1987, for reviews). The focus is on processes and abilities and
the theoretical result is a sequence of steps or stages in the development of
aspects of cognition. This perspective has been fruitful for describing and
explaining the development of ethnic identity, but there are also challenges.
For example, it is unclear to what extent these models are applicable
to children from ethnic minority groups (R. Brown, 1995; Coll et al.,
1996). Furthermore, general developments in the structural complexity of
children’s thinking and the way children process information and acquire
knowledge do not say much about the particular beliefs, knowledge, and
attitudes children acquire or express. This depends on what is available
and common in children’s social environments. Cognition is embedded
in historical, cultural, and sociorelational contexts (see Resnick, Levine, &
Teasley, 1991; Thompson & Fine, 1999). Ethnic identity does not only
depend on cognitive structures and processes but is also determined by
social beliefs and context. Although there are important exceptions (e.g.,
Bronfenbrenner, 1979; Emler, 1987), developmental psychology has typically
elaborated the internal, psychological aspects of development. The
emphasis is on psychological rather than social processes. However, as I
have tried to show, social information about status differences, cultural
values, and situational conditions play an important role in children’s
ethnic self-understanding.
In social psychology, SIT and particularly SCT emphasise the importance

offered by Hong et al. (2000), who present a more dynamic constructivist rather than struc-
tural model for understanding what they call frame switching in bicultural individuals. They
argue that culture is not internalised in the form of an integrated structure but rather as
domain-specific knowledge, such as implicit theories. Furthermore, individuals would be able
to acquire more than one cultural frame, but these frames will not simultaneously guide
thinking. Accessibility of constructs and contextual salience are the concepts used by Hong et
al. to explain how particular pieces of cultural knowledge become operative in particular
situations. These ideas are quite similar to SCT and offer the possibility to connect SCT with
theories developed in cross-cultural work.
208 Ethnic identity and social context
of the social context. These theories are very useful for examining how
children make sense of ethnic differences in particular social settings. Ethnic
self-categorisation is considered to be a function of the relative size and status
of the group, and the cultural and normative meanings associated with par-
ticular social categories in interaction with the child’s readiness to make eth-
nic distinctions. This chapter has used these ideas for presenting some of the
results of our studies on ethnic minority children in the Netherlands. The
separate studies can be interpreted in different ways, but the findings as a
whole can be understood in terms of these social-psychological theories. The
research supports some of the key ideas of these theories, but also raises
additional issues and questions.
For example, perceived discrimination and collectivism were examined as
two central conditions for the development of ethnic identity among minor-
ity groups. Both have an effect on children’s ethnic identification. This shows
that not only the “minority” or structural aspect of ethnic minorities – which
is stressed by SIT – but also the “ethnic” or cultural aspect needs to be
considered.
There are also limitations to the way intergroup theories are typically pre-
sented and used. In many countries around the world, ethnic group situations
consist of a series of groups that differ in cultural background and in which
ethnicity or related characteristics such as race, language, and religion are
criteria for group differences. Studies drawing upon SIT have a tendency,
however, to focus exclusively on status differences between the majority and
minority groups. A majority–minority dichotomy is typically used, in which
there is a tendency to treat minority groups as a homogeneous category.
Differences between ethnic minority groups and within-group differences are
often not considered.
Obviously, there are important differences between the group of majority
children and that of minority children. For example, minorities perceive,
in general, more personal and group discrimination than majority group
children do. Hence, minority group children seem to be well aware of their
group’s lower status and do not deny the existence of personal discrimin-
ation. Not all minority groups, however, are perceived in the same way.
Studies in various countries have found that different minority groups
experience varying degrees of social acceptability (see Hagendoorn, 1995;
Owen, Eisner, & McFaul, 1981, for reviews). For example, in the Netherlands,
studies on children, adolescents, and adults have found that the Turks are the
least liked minority group (Hagendoorn, 1995; Verkuyten & Kinket, 2000).
In agreement with this finding, Turkish children have been found to perceive
more personal and group discrimination than, for example, Moroccan
and Surinamese children. Furthermore, the negative relationship between
discrimination and ethnic self-esteem is most evident among the Turks. In
addition, there are also many visible and cultural features that shape the
everyday life and life experiences of ethnic minority groups differently.
Turkish children have been found, for example, to have a more collectivist
 Verkuyten 209
value orientation compared to Surinamese children. Also, related to their
Islamic background, gender differentiation is more evident on the level of
beliefs and actual practices among the Turks and Moroccans than among, for
example, the Surinamese. In fact, among the two Islamic groups, there is a
strong association between gender and ethnicity. Turkish and Moroccan
people in the Netherlands partly define their identity in terms of their
cultural specific beliefs about gender roles.
Thus, for understanding the development of ethnic identity, it is not only
differences between the majority group and the minorities as a group that are
relevant, but also differences between ethnic minority groups. In addition,
there are individual differences – such as experiences with discrimination
and in collectivism – that highlight the need to examine distinctions among
children in general and within the same ethnic group in particular. Attention
to between-group differences should not lead to overlooking within-group
heterogeneity (Celious & Oyserman, 2001). Hence, to understand the devel-
opment of ethnic identity it is necessary to pay attention to within-group
differences, in addition to differences between ethnic minority groups,
between the majority group and ethnic minorities, and also to more basic
capacities and tendencies that pervade across ethnic lines. Furthermore, there
are not only ethnic distinctions but also differences associated with other
important characteristics, such as gender (Celious & Oyserman, 2001;
Frable, 1997). In addition, ethnic identity is a multidimensional construct.
Its various components may develop differently and be related differently to
self-perceptions (e.g., Garza & Herringer, 1987; Phinney, 1990).
SIT and SCT emphasise the role of social context but say little about the
development of ethnic identity. These theories do not systematically address
developmental issues. SIT focuses on processes of social comparison and self-
evaluation, and in SCT cognitive-developmental constraints are typically not
considered. For SCT, development seems to be a question of growing flexibil-
ity in categorical judgments and increasing understanding and knowledge of
the social meanings associated with social categories. However, the particular
cognitive and affective correlates of children’s social understandings and
of their proclivity to ethnic self-definitions are not addressed. The emphasis
is on the processes of category use in context, and not on how self-
understandings are conceptually organised and change in structure with age9.
Cognitive-developmental theories make a contribution here because they try
to account for qualitative reorganisations in children’s ideas and feelings
about themselves (Corenblum & Annis, 1993). Hence, cognitive theories are
clearly important for understanding ethnic identity development. The same is

9 However, SCT presents a dynamic and functional approach to the self. The focus is on com-
parative, relational judgments that define people in their social context; that is, on self-
conception rather than on a structured self-concept. Turner and Onorato (1999) have rejected
structural models of the self for reasons of cognitive economy and for being unable to explain
the formation of new categories (Turner, 1999).
210 Ethnic identity and social context
true, for example, for theories that focus on social learning processes. How-
ever, as I have tried to argue, social-psychological intergroup theories offer a
useful additional framework for examining ethnic identity in context. A focus
on context provides an important perspective on the multifaceted question
about how children learn to adapt to their ethnically diverse world, and the
complex issue of ethnic self-understanding.

REFERENCES

Aboud, F. E. (1987). The development of ethnic self-identification and attitudes. In

J. S. Phinney & M. J. Rotheram (Eds.), Children’s ethnic socialization: Pluralism and
development (pp. 32–55). Newbury Park, CA: SAGE Publications.
Aboud, F. (1988). Children and prejudice. Oxford: Blackwell.
Bagley, C., & Coard, B. (1975). Cultural knowledge and rejection of ethnic identity in
West Indian children in London. In G. K. Verma & C. Bagley (Eds.), Race and
education across cultures (pp. 321–331). London: Heinemann.
Banks, W. C. (1976). White preference in blacks: A paradigm in search of a
phenomenon. Psychological Bulletin, 83, 1179–1186.
Bennett, M., Sani, F., Lyons, E., & Barrett, M. (1998). Children’s subjective
identification with the group and in-group favouritism. Developmental Psychology,
34, 902–909.
Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorization and the
formation of intergroup attitudes in children. Child Development, 68, 530–543.
Brand, E. S., Ruiz, R. A., & Padilla, A. M. (1974). Ethnic identification and
preference: A review. Psychological Bulletin, 81, 860–890.
Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999). Perceiving
pervasive discrimination among African Americans: Implications for group
identification and well-being. Journal of Personality and Social Psychology, 77,
135–149.
Brewer, M. B. (1999). The psychology of prejudice: In-group love or out-group hate?
Journal of Social Issues, 55, 429–444.
Brewer, M. B., & Miller, N. (1996). Intergroup relations. Buckingham, UK: Open
University Press.
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature
and design. Cambridge, MA: Harvard University Press.
Brown, L. M. (1998). Ethnic stigma as a contextual experience: A possible selves
perspective. Personality and Social Psychology Bulletin, 24, 163–172.
Brown, R. (1995). Prejudice: Its social psychology. Oxford: Blackwell.
Brown, R. (2000). Social identity theory: Past achievements, current problems and
future challenges. European Journal of Social Psychology, 30, 745–778.
Brown, R., & Williams, J. A. (1984). Group identification: The same thing to all
people. Human Relations, 37, 547–564.
Cameron, J. A., Alvarez, J. M., Ruble, D. N., & Fuligni, A. J. (2001). Children’s lay
theories about in-groups and out-groups: Reconceptualizing research on prejudice.
Personality and Social Psychology Review, 5, 118–128.
Celious, A., & Oyserman, D. (2001). Race from the inside: An emerging
heterogeneous race model. Journal of Social Issues, 57, 149–165.
 Verkuyten 211
Clark, K. B., & Clark, M. P. (1947). Racial identification and preference in Negro
children. In T. M. Newcombe & E. L. Hartley (Eds.), Readings in social psychology
(pp. 169–178). New York: Holt, Rinehart & Winston.
Coll, C. G., Lamberty, G., Jenkins, R., McAdoo, H. P., Crnic, K., Wasik, B. H., &
Garcia, H. V. (1996). An integrative model for the study of developmental
competencies in minority children. Child Development, 67, 1891–1914.
Corenblum, B., & Annis, R. C. (1993). Development of racial identity in minority and
majority children: An affect discrepancy model. Canadian Journal of Behavioral
Science, 25, 499–521.
Cornell, S., & Hartmann, D. (1998). Ethnicity and race: Making identities in a
changing world. London: Pine Forge Press.
Cousins, S. D. (1989). Culture and self-perception in Japan and the United States.
Journal of Personality and Social Psychology, 56, 124–131.
Crocker, J., & Luhtanen, R. (1990). Collective self-esteem and in-group bias. Journal
of Personality and Social Psychology, 58, 60–67.
Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. T. Gilbert, S. T. Fiske,
& G. Lindzey (Eds.), The handbook of social psychology (4th ed., pp. 504–553). New
York: McGraw-Hill.
Cross, W. E. (1991). Shades of Black: Diversity in African-American identity.
Philadelphia, PA: Temple University Press.
Davey, A. G., & Mullin, P. N. (1980). Ethnic identification and preference of
British primary school children. Journal of Child Psychology and Psychiatry, 21,
241–251.
DeVos, G. A. (1995). Ethnic pluralism, conflict and accommodation. In L. Romanucci-
Ross & G. DeVos (Eds.), Ethnic identity: Creation, conflict and accommodation
(pp. 15–47). Walnut Creek, CA: AltaMira.
DeVries, M. (1987). Ogen in je rug. Alphen aan den Rijn, The Netherlands:
Samsom.
Ellemers, N., Kortekaas, P., & Ouwekerk, J. W. (1999). Self-categorisation, commit-
ment to the group and group self-esteem as related but distinct aspects of social
identity. European Journal of Social Psychology, 29, 371–189.
Emler, N. (1987). Socio-moral development from the perspective of social representa-
tions. Journal for the Theory of Social Behaviour, 17, 371–388.
Emler, N., & Ohana, J. (1993). Studying social representations in children. In G. M.
Breakwell & D. V. Canter (Eds.), Empirical approaches to social representations
(pp. 63–89). Oxford: Clarendon.
Fine, M., & Bowers, C. (1984). Racial self-identification: The effects of social history
and gender. Journal of Applied Social Psychology, 14, 136–146.
Fishbein, H. D. (1996). Peer prejudice and discrimination: Evolutionary, cultural, and
developmental dynamics. Boulder, CO: Westview Press.
Frable, D. E. S. (1997). Gender, racial, ethnic, sexual, and class identities. Annual
Review of Psychology, 48, 139–162.
Gaines, S. O., Marelich, W. D., Bledsoe, K. L., Steers, W. N., Henderson, M. C.,
Granrose, C. S., Barájas, L., Hicks, D., Lyde, M., Takahashi, Y., Yum, N., Ríos,
D. I., Garciá, B. F., Farris, K. R., & Page, M. S. (1997). Links between race/ethnicity
and cultural values as mediated by racial/ethnic identity and moderated by gender.
Journal of Personality and Social Psychology, 72, 1460–1476.
Garza, R. T., & Herringer, L. G. (1987). Social identity: A multidimensional
approach. The Journal of Social Psychology, 127, 299–308.
212 Ethnic identity and social context
Goodman, M. E. (1952). Race awareness in young children. New York: Macmillan.
Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: Attitudes,
self-esteem, and stereotypes. Psychological Review, 102, 4–27.
Hagendoorn, L. (1995). Intergroup biases in multiple group systems: The perception
of ethnic hierarchies. In W. Stroebe & M. Hewstone (Eds.), European review of
social psychology (pp. 199–228). Chichester, UK: John Wiley & Sons.
Harter, S. (1999). The construction of the self: A developmental perspective. New York:
Guilford Press.
Hawker, D. S. J., & Boulton, M. J. (2000). Twenty years’ research on peer victimization
and psychosocial maladjustment: A meta-analytic review of cross-sectional studies.
Journal of Child Psychology and Psychiatry, 41, 441–455.
Hetts, J. J., Sakuma, M., & Pelham, B. W. (1999). Two roads to positive regard:
Implicit and explicit self-evaluation and culture. Journal of Experimental Social
Psychology, 35, 512–559.
Hofstede, G. (1980). Culture’s consequences. Beverly Hills, CA: SAGE Publications.
Hong, Y.-Y., Morris, M. W., Chiu, C. Y., & Benet-Martinez, V. (2000). Multicultural
minds: A dynamic constructivist approach to culture and cognition. American
Psychologist, 55, 709–720.
Horowitz, D. L. (1985). Ethnic groups in conflict. Berkeley, CA: University of
California Press.
Horowitz, E. L. (1939). Racial aspects of self-identification in nursery school children.
Journal of Psychology, 7, 91–99.
Hraba, J., & Grant, G. (1970). Black is beautiful: A re-examination of racial prefer-
ence and identification. Journal of Personality and Social Psychology, 16, 398–402.
Huiberts, A. M., Vollebergh, W. A. M., & Meeus, W. (1999). Individualisme en
collectivisme bij Nederlandse, Turkse en Marokkaanse jongeren. Tijdschrift voor
Orthopedagogiek, 38, 342–356.
Hutnik, N. (1991). Ethnic minority identity. Oxford: Clarendon.
Jackson, J. W., & Smith, E. R. (1999). Conceptualizing social identity: A new frame-
work and evidence for the impact of different dimensions. Personality and Social
Psychology Bulletin, 25, 120–135.
Kaǧitçibaşi, C. (1990). Family and socialisation in cross-cultural perspective: A
model of change. In J. Berman (Ed.), Nebraska symposium on motivation 1989:
Cross-cultural perspectives. Lincoln: University of Bebraska Press.
Kardiner, A., & Ovesey, L. (1951). The mark of oppression. New York: Norton.
Keefe, S. E. (1992). Ethnic identity: The domain of perceptions and attachment to
ethnic groups and cultures. Human Organization, 51, 35–43.
Kenny, D. A., & Judd, C. M. (1984). Estimating the nonlinear and interactive effects
of latent variables. Psychological Bulletin, 96, 201–210.
Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social psychology. In
D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology
(4th ed., pp. 233–265). Boston, MA: McGraw-Hill.
Kinket, B., & Verkuyten, M. (1997). Levels of ethnic self-identification and social
context. Social Psychology Quarterly, 60, 338–354.
LaFromboise, T., Coleman, H. L. K., & Gerton, J. (1993). Psychological impact of
biculturalism: Evidence and theory. Psychological Bulletin, 114, 395–412.
Lay, C., Fairlie, P., Jackson, S., Ricci, T., Eisenberg, J., Sato, T., Teeäär, A., &
Melamud, A. (1998). Domain-specific allocentrism–idiocentrism: A measure of
family connectedness. Journal of Cross-Cultural Psychology, 29, 434–460.
 Verkuyten 213
Leary, M. R., & Baumeister, R. F. (2000). The nature and function of self-esteem:
Sociometer theory. In M. Zanna (Ed.), Advances in experimental social psychology,
Vol. 32 (pp. 1–61). San Diego, CA: Academic Press.
Liebkind, K. (1992). Ethnic identity: Challenging the boundaries of social psy-
chology. In G. M. Breakwell (Ed.), Social psychology of identity and the self-concept.
London: Surrey University Press.
Mahan, J. (1976). Black and white children’s racial identification and preference.
Journal of Black Psychology, 3, 47–53.
Markus, H. R., Kitayama, S., & Heiman, R. J. (1999). Culture and “basic” psycho-
logical principles. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology:
Handbook of basic principles (pp. 857–913). New York: Guilford Press.
Milner, D. (1973). Racial identification and preference in “black” British children.
European Journal of Social Psychology, 3, 281–295.
Milner, D. (1983). Children and race: Ten years on. London: Ward Lock Educational.
Morland, J. K. (1969). Race awareness among American and Hong Kong Chinese
children. American Journal of Sociology, 75, 360–374.
Moscovici, S. (1984). The phenomenon of social representations. In R. Farr & S.
Moscovici (Eds.), Social representations (pp. 3–69). Cambridge: Cambridge Uni-
versity Press.
Nesdale, D. (2000). Developmental changes in children’s ethnic preferences and social
cognitions. Journal of Applied Developmental Psychology, 20, 501–519.
Nesdale, D. (2001). The development of prejudice in children. In M. Augoustinos
& K. J. Reynolds (Eds.), Understanding prejudice, racism and social conflict
(pp. 57–72). London: SAGE Publications.
Nijsten, C. (1998). Opvoeden in Turkse gezinnen in Nederland. Assen, The
Netherlands: Van Gorcum.
Oakes, P. J., Haslam, S. A., & Turner, J. C. (1994). Stereotyping and social reality.
Oxford: Blackwell.
Owen, C., Eisner, H., & McFaul, T. (1981). A half-century of social distance research:
National replication of the Bogardus studies. Sociology and Social Research, 66,
80–98.
Oyserman, D. (1993). The lens of personhood: Viewing the self and others in a
multicultural society. Journal of Personality and Social Psychology, 65, 993–1009.
Oyserman, D., & Swim, J. K. (2001). Stigma: An insider’s perspective. Journal of
Social Issues, 57, 1–14.
Pelham, B. W., & Hetts, J. J. (1999). Implicit and explicit personal and social identity:
Toward a more complete understanding of the social self. In T. R. Tyler, R. M.
Kramer., & O. P. John (Eds.), The psychology of the social self (pp. 115–143).
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Phinney, J. (1989). Stages of ethnic identity in minority group adolescents. Journal of
Early Adolescents, 9, 34–49.
Phinney, J. S. (1990). Ethnic identity in adolescents and adults: A review of research.
Psychological Bulletin, 108, 499–514.
Phinney, J. S., Madden, T., & Santos, L. J. (1998). Psychological variables as predictors
of perceived ethnic discrimination among minority and immigrant adolescents.
Journal of Applied Social Psychology, 28, 937–953.
Phinney, J. S., & Rosenthal, D. A. (1992). Ethnic identity in adolescence: Process,
context, and outcome. In G. R. Adams, T. P. Gullotta, & R. Montemayor (Eds.),
Adolescent identity formation (pp. 145–172). New York: Academic Press.
214 Ethnic identity and social context
Phinney, J. S., & Rotheram, M. J. (Eds.), Children’s ethnic socialization: Pluralism and
development. Newbury Park, CA: SAGE Publications.
Powlishta, K. K., Serbin, L. A., Doyle, A.-B., & White, D. R. (1994). Gender,
ethnicity, and body type biases: The generality of prejudice in childhood.
Developmental Psychology, 30, 526–536.
Ramsey, P. G. (1987). Young children’s thinking about ethnic differences. In J. S.
Phinney & M. J. Rotheram (Eds.), Children’s ethnic socialization: Pluralism and
development (pp. 56–72). Newbury Park, CA: SAGE Publications.
Resnick, L. B., Levine, J. M., & Teasley, S. D. (Eds.) (1991). Perspectives on socially
shared cognition. Washington, DC: American Psychological Association.
Rhee, E., Uleman, J. S., Lee, H. K., & Roman, R. J. (1995). Spontaneous
self-descriptions and ethnic identities in individualistic and collectivistic cultures.
Journal of Personality and Social Psychology, 69, 142–152.
Roosens, E. (1994). The primordial nature of origins in migrant ethnicity. In H.
Vermeulen & C. Govers (Eds.), The anthropology of ethnicity: Beyond “ethnic
groups and boundaries”. Amsterdam: Spinhuis.
Rosenberg, M. (1979). Conceiving the self. New York: Basic Books.
Rotenberg, K. J., & Cranwell, F. R. (1989). Self-concept in American Indian and
White children. Journal of Cross-Cultural Psychology, 20, 39–53.
Rotheram, M. J., & Phinney, J. S. (1987). Introduction: Definitions and perspectives in
the study of children’s ethnic socialization. In J. S. Phinney & M. J. Rotheram
(Eds.), Children’s ethnic socialization: Pluralism and development (pp. 10–28).
Newbury Park, CA: SAGE Publications.
Rutland, A. (1999). The development of national prejudice, in-group favouritism and
self-stereotypes in British children. British Journal of Social Psychology, 38, 55–70.
Sellers, R. M., Smith, M. A., Shelton, J. N., Rowley, S. A. J., & Chavous, T. M. (1998).
Multidimensional model of racial identity: A reconceptualization of African
American racial identity. Personality and Social Psychology Review, 2, 18–39.
Singelis, T. M. (1994). The measurement of independent and interdependent
self-construals. Personality and Social Psychology Bulletin, 20, 580–591.
Smith, P. B., & Bond, M. H. (1993). Social psychology across cultures: Analysis and
perspectives. New York: Harvester Wheatsheaf.
Smith, E. R., Murphy, J., & Coats, S. (1999). Attachments to groups: Theory and
measurement. Journal of Personality and Social Psychology, 77, 94–110.
Spielman, D. A. (2000). Young children, minimal groups, and dichotomous
categorization. Personality and Social Psychology Bulletin, 26, 1433–1441.
Swim, J. K., & Stangor, C. (1998). Prejudice: The target’s perspective. New York:
Academic Press.
Tajfel, H. (1978). The social psychology of minorities. London: Minority Rights
Group.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior.
In S. Worchel & W. Austin (Eds.), Psychology of intergroup relations (pp. 7–24).
Chicago, IL: Nelson-Hall.
Taylor, D. M., Wright, S. C., Moghaddam, F. M., & Lalonde, R. N. (1990). The
personal/group discrimination discrepancy: Perceiving my group, but not myself,
to be a target of discrimination. Personality and Social Psychology Bulletin, 16,
254–262.
Taylor, D. M.,Wright, S. C., & Porter, L. E. (1993). Dimensions of perceived dis-
crimination: The personal/group discrimination discrepancy. In M. P. Zanna &
 Verkuyten 215
J. M. Olson (Eds.), The psychology of prejudice: The Ontario symposium, Vol. 7
(pp. 233–255). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Thompson, L., & Fine, G. A. (1999). Socially shared cognition, affect, and
behavior: A review and integration. Personality and Social Psychology Review, 3,
278–302.
Trafimow, D., Triandis, H. C., & Goto, S. G. (1991). Some tests of the distinction
between the private and the collective self. Journal of Personality and Social
Psychology, 60, 649–655.
Triandis, H. C. (1994). Theoretical and methodological approaches to the study of
collectivism and individualism. In U. Kim, H. C. Triandis, C. Kagitcibasi,
S. C. Choi, & G. Yoon (Eds.), Individualism and collectivism: Theory, method, and
applications (pp. 41–51). Thousand Oaks, CA: SAGE Publications.
Triandis, H. C., Leung, K., Villareal, M. J., & Clack, F. L. (1985). Allocentric and
idiocentric tendencies: Convergent and discriminant validation. Journal of Research
in Personality, 19, 395–415.
Turner, J. C. (1999). Some current issues in research on social identity and
self-categorization theories. In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social
identity: Context, commitment, content (pp. 6–34). Oxford: Blackwell.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorization theory. Oxford: Blackwell.
Turner, J. C., Oakes, P. J., Haslam, S. A., & McGarty, C. (1994). Self and collective:
Cognition and social context. Personality and Social Psychology Bulletin, 20,
454–463.
Turner, J. C., & Onorato, R. S. (1999). Social identity, personality, and the self-
concept: A self-categorization perspective. In T. R. Tyler, R. M. Kramer, & O. P.
John (Eds.), The psychology of the social self. Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Van den Heuvel, H., Tellegen, G., & Koomen, W. (1992). Cultural differences in the
use of psychological and social characteristics in children’s self-understanding.
European Journal of Social Psychology, 22, 353–362.
Vaughan, G. M. (1964). The development of ethnic attitudes in New Zealand school
children. Genetic Psychology Monographs, 70, 135–175.
Vaughan, G. M. (1987). A social psychological model of ethnic identity development.
In J. S. Phinney & M. J. Rotheram (Eds.), Children’s ethnic socialization: Pluralism
and development (pp. 73–91). Newbury Park, CA: SAGE Publications.
Verkuyten, M. (1999). Etnische identiteit: Theoretische en empirische benaderingen.
Amsterdam: Spinhuis.
Verkuyten, M. (2001). Global self-esteem, ethnic self-esteem and family integrity:
Turkish and Dutch early adolescents in The Netherlands. International Journal of
Behavioural Development, 25, 357–366.
Verkuyten, M. (2002a). Ethnic attitudes among minority and majority children:
The role of ethnic identification, peer group victimisation and parents. Social
Development, 11, 558–570.
Verkuyten, M. (2002b). Perceptions of ethnic discrimination by minority and
majority early adolescents in The Netherlands. International Journal of Psychology,
37, 321–332.
Verkuyten, M., & Kinket, B. (2000). Social distances in a multi ethnic society:
The ethnic hierarchy among Dutch preadolescents. Social Psychology Quarterly, 63,
75–85.
216 Ethnic identity and social context
Verkuyten, M., Kinket, B., & Van der Wielen, C. (1997). Preadolescents’
understanding of ethnic discrimination. Journal of Genetic Psychology, 158,
97–112.
Verkuyten, M., & Masson, K. (1996). Culture and gender differences in the perception
of friendship by adolescents. International Journal of Psychology, 31, 207–217.
Verkuyten, M., & Pouliasi, K. (2002). Biculturalism among older children:
Cultural frame-switching, attributions, self-identification and attitudes. Journal of
Cross-Cultural Psychology, 33, 596–608.
Verkuyten, M., & Thijs, J. (2000). Leren (en) waarderen: Discriminatie, zelfbeeld,
relaties en leerprestaties in ‘witte’en ‘zwarte’ basisscholen. Amsterdam: Thela Thesis.
Verkuyten, M., & Thijs, J. (2001). Peer victimization and self-esteem of ethnic
minority group children. Journal of Community and Applied Social Psychology, 11,
227–234.
Verkuyten, M., & Thijs, J. (2002). Racist victimization among children in the
Netherlands: The effect of ethnic group and school. Ethnic and Racial Studies, 25,
310–331.
Verkuyten, M., & Thijs, J. (2003). Global and ethnic self-esteem in social context:
Minority and majority groups in the Netherlands. Social Indicators Research
(forthcoming).
Weber, M. (1968). Economy and society: An outline of interpretive sociology. New
York: Bedmeister.
Weinreich, P. (1986). The operationalisation of identity theory in racial and ethnic
relations. In J. Rex & D. Mason (Eds.), Theories of race and ethnic relations
(pp. 299–344). Cambridge: Cambridge University Press.
Yee, M., & Brown, R. (1992). Self evaluations and intergroup attitudes in children
aged three to nine. Child Development, 63, 619–629.
Part III

Applications
8 Social identity processes and
children’s ethnic prejudice
Drew Nesdale

Although the presence of ethnic prejudice is destructive in any sector of a

community, the possibility that it may emerge in school-age children is of
particular concern. During this period, children acquire social knowledge
and attitudes that often endure into adulthood and which, in the case of
ethnic prejudice, would foster intergroup divisions and, at worst, physical
harm to members of minority groups.
In response to these long-held concerns, a considerable amount of research
has addressed the development of children’s ethnic prejudice (i.e., their
feelings of dislike or hatred towards members of ethnic out-groups), to-
gether with related issues such as the acquisition of ethnic awareness, ethnic
self-identification, and ethnic stereotyping (see reviews by Aboud, 1988;
D. Nesdale, 2001a). Given the difficulties in measuring such constructs,
especially in young children, researchers have used a variety of measurement
techniques, including ethnic preferences, trait assignments, structured inter-
views, behaviour observations, questionnaires, sociograms, and projective
tests, although the first two, ethnic preferences and trait assignments, have
dominated the field (see D. Nesdale, 2001a, for a review).
Based largely on the ethnic preference and trait attribution techniques, a
remarkably consistent set of findings has emerged, especially in relation to
dominant group children. The ethnic preference studies indicate that children
can differentiate among people based on racial cues (e.g., skin colour) from
a very early age and certainly by around 4 years of age their racial awareness
enables them to distinguish explicitly among members of different racial
groups. There is also extensive evidence that from 4 years of age onwards,
children from the ethnically dominant group can accurately identify their
own ethnic group membership, and that they reveal increasingly strong
in-group bias in their choices. Similarly, trait attribution studies have con-
sistently revealed that dominant group children display an increase in in-
group positivity/out-group negativity in their trait attributions from 3 to 4
years of age (see reviews by Aboud, 1988; D. Nesdale, 2001a). Based on her
review of the literature and her own more recent findings, Aboud further
concluded that this bias actually peaks at around 6 to 7 years of age, and then
gradually declines during mid childhood.
220 Social identity and ethnic prejudice
However, despite the apparently high level of consistency in the findings,
an agreed-upon understanding of the nature of the development of ethnic
prejudice in children has yet to emerge. A major reason relates to the nature
of the foregoing findings. Although some researchers argue to the contrary
(e.g., Aboud, 1988; Gregor & McPherson, 1966), a common view of the
preceding findings is that they simply reflect the children’s in-group prefer-
ences rather than their feelings of out-group dislike or hatred (e.g., Brand,
Ruiz, & Padilla, 1974; Cameron, Alvarez, Ruble, Fuligni, 2001; Katz, 1976;
D. Nesdale, 2001a; Proshansky, 1966; Stephan & Rosenfield, 1979). On this
basis alone, it is not surprising that there is currently little agreement on issues
such as the age at which prejudice emerges in children, whether or not there
are age-related phases or stages through which prejudice develops, what the
psychological processes or mechanisms are which govern the acquisition of
ethnic prejudice, and what impact is exerted by children’s emerging linguistic
and cognitive abilities upon their acquisition and retention of ethnic
prejudice.
The aim of this chapter is to outline a model that is designed to provide
a comprehensive account of the development of children’s prejudice. The
model draws heavily upon social identity theory (SIT; Tajfel & Turner, 1979)
and its more recent elaboration, self-categorisation theory (SCT; Turner,
Hogg, Oakes, Reicher, & Wetherell, 1987), which arguably provides the most
widely endorsed current explanation of ethnic prejudice in adults. Since SIT
(and SCT) was not designed to explain the development of prejudice in chil-
dren, social identity development theory (SIDT) is proposed to account for
the acquisition of prejudice in children, taking into account their unique
circumstances of change and development as they increase in age. Central
to SIDT is the assumption that social identity processes fundamentally
influence the development of children’s prejudice.
To provide a context for the description of SIDT, the three main explan-
ations of children’s prejudice that have been proposed to date will be briefly
considered. These approaches include the view that ethnic prejudice in chil-
dren is a form of emotional maladjustment; that ethnic prejudice emerges in
children as a reflection of the views of significant others; and that ethnic
prejudice is determined by changes in children’s perceptual-cognitive acquisi-
tions. A description of the main tenets of SIDT follows, together with the
results of recent studies designed to assess the theory.

EXPLANATIONS OF CHILDREN’S PREJUDICE

The emotional maladjustment approach links the acquisition of prejudice in

children to the development of a particular personality type, the authoritar-
ian personality (Adorno, Frenkel-Brunswik, Levinson, & Sanford, 1950).
Much of children’s prejudice influenced by Freudian thinking, is considered
to stem from emotional maladjustment arising from a repressive and harshly
 Nesdale 221
disciplined upbringing. Under these circumstances, the child’s resulting frus-
tration, anger, and hostility towards his/her parents is considered to be dis-
placed away from the parents towards scapegoats who are weaker and lack
authority and power, such as members of minority groups (D. Nesdale,
1999a).
While a positive feature of this theory is that it provides an explanation for
any differences in levels of prejudice that may occur between individuals
(Aboud, 1988), it does not account for the uniformity of prejudice across
whole groups of people that may occur in particular places and times, nor
why some groups are the recipients of prejudice but not others (Brown, 1995).
In general, the approach ignores the influence on people’s (including child-
ren’s) intergroup attitudes and behaviour of important aspects of their social
environment, including the attitudes of significant others, prevailing societal
norms, and the relationships between the members of the dominant cultural
group and the members of minority groups.
The social reflection approach takes the latter point as a fundamental prem-
ise – children’s prejudice is considered to reflect the community’s attitudes
and values, which are typically transmitted by the child’s parents. The most
widely accepted version of this approach has been that children simply learn
their ethnic attitudes from these sources in the same way that they learn other
social behaviours (e.g., Allport, 1954; Rosenfield & Stephan, 1981). Thus, as
Horowitz put it as far back as 1936, “attitudes toward Negroes, are now
chiefly determined not by contact with Negroes but by contact with the
prevalent attitude towards Negroes” (pp. 34–35). Presumably, such learning
occurs because the children are rewarded for their imitative behaviour and/or
want to please their parents.
However, although the importance accorded to parents (and, later, peers)
as the primary source of children’s ethnic attitudes has remained largely
unchallenged over many years, research support for this approach is actually
quite mixed. On the one hand, consistent with it are findings of positive
correlations between the ethnic attitudes of children and their parents (e.g.,
Bird, Monachesi, & Burdick, 1952; Goodman, 1952; Harris, Gough, &
Martin, 1950; Horowitz & Horowitz, 1938; Mosher & Scodel, 1960; Radke
& Trager, 1950) and that there are distinct similarities in the statements
of parents and their children concerning ethnic minority groups (e.g.,
Radke-Yarrow, Trager, & Miller, 1952). On the other hand, however, some
studies have reported either low (e.g., Bird et al., 1952; Frenkel-Brunswik
& Havel, 1953) or nonexistent correlations (e.g., Aboud & Doyle, 1996b;
Pushkin, cited in Davey, 1983) between children and their parents in terms
of their ethnic attitudes. Still other studies have found that when there is a
similarity between parents and children in their negative statements towards
ethnic minorities, the children frequently explicitly source the statements to
their parents, declining to take ownership of them (e.g., Horowitz & Horowitz,
1938; Porter, 1971; Radke, Trager, & Davis, 1949). In addition, although
children as young as 5 years of age incorporated trait-like terms or implied
222 Social identity and ethnic prejudice
dispositions (e.g., “bad”, “dirty”, “ugly”) in their verbal descriptions of
minority group children (e.g., Horowitz & Horowitz, 1938; Porter, 1971;
Radke-Yarrow et al., 1952), there was little sense that such terms had trait
connotations for them and/or that they were part of a set of beliefs that were
shared by their group (i.e., a stereotype). In other words, there was scant
evidence that the attributes or behaviours were perceived to reflect stable
characteristics of members of the particular ethnic out-group that might be
revealed on other occasions (see D. Nesdale, 2001b, for a review of the rela-
tionship between parents, language, and ethnic prejudice in children).
Overall, the preceding discussion suggests that it would be incorrect to
assume that children should be regarded as empty containers into which
prevailing societal prejudices are poured, or as sponges that soak up domin-
ant ethnic attitudes (R. Brown, 1995; Davey, 1983; Milner, 1996). Children
may hold the same ethnic attitudes as their parents and peers, but this situ-
ation does not arise as a matter of necessity. As a considerable amount of
research has made clear, children’s intellectual capacities reveal dramatic
development through the mid childhood years and they are active partici-
pants in seeking to understand and control both their cognitive and social
worlds (e.g., Durkin, 1995).
To a certain extent, the latter view accords with recent accounts of
children’s prejudice that have emphasised the influence exerted on children’s
ethnic attitudes by perceptual-cognitive considerations (e.g., Aboud, 1988;
Aboud & Doyle, 1996a, 1996b; Bigler, 1995).
Aboud and her colleagues, for example, have argued that there is no strong
evidence that children’s prejudice is influenced by the attitudes of parents and
peers. Instead, according to Aboud’s (1988) sociocognitive theory (ST),
a child’s attitude to other groups of children depends on his/her levels of
development in relation to two overlapping sequences of perceptual-cognitive
development. One sequence involves the process that dominates a child’s
experience at a particular time. The child is initially dominated by affective-
perceptual processes associated with fear of the unknown and attachment to
the familiar. Perceptual processes subsequently dominate, preference for the
(similar) in-group and rejection of the (different) out-group being determined
primarily by physical attributes (e.g., skin colour, language, body size).
Thereafter, cognitive processes take ascendancy with the advent of the con-
crete operational stage of cognitive development around 7 years of age and,
later, formal operational thinking (Flavell, 1963). The effect of the transition
to cognitive processes is that the child is increasingly able to understand
the individual rather than the group-based qualities of people. Overlapping
this sequence is a second sequence of development concerned with changes in
the child’s focus of attention. Whereas very young children mostly focus on
themselves and their preferences and perceptions, older children emphasize
categories of people such that individuals are seen as members of these
categories or groups. Still later, however, children focus on individuals, who
are liked or disliked for their personal rather than group qualities.
 Nesdale 223
Based on these sociocognitive developments, Aboud (1988) argued that
in-group bias and out-group prejudice increase to a peak between 5 to 7 years
of age, when group differences are paramount. However, with subsequent
increases in the child’s cognitive abilities, occasioned by the onset of concrete
operational thinking around 7 years of age, Aboud claimed that there is a
systematic decline in group-based biases, which is further enhanced when
the child’s ever-increasing cognitive abilities allow him/her to attend to the
differences between individuals.
Although ST offers a theoretically consistent account of children’s preju-
dice in terms of their perceptual-cognitive processes, it is fair to say that
research support for this theory is also currently mixed (see D. Nesdale,
2001a, 2001b, for a review). On the one hand, consistent with ST are studies
in which children from the ethnically dominant group displayed increasing
in-group positivity/out-group negativity up to 6 to 7 years of age, followed by
a systematic decline (e.g., Aboud, 1988; Bigler & Liben, 1993; Doyle &
Aboud, 1995). In addition, research has revealed that children’s understand-
ing of conservation (an achievement of the concrete operational stage of
cognitive development) is correlated with ethnic flexibility, the understanding
that ethnically similar and different individuals can have different and similar
attributes, respectively (e.g., Doyle, Beaudet, & Aboud, 1988), and ethnic
constancy, the understanding that ethnicity remains the same despite super-
ficial transformations in skin colour or clothing (e.g., Aboud, 1984; Semaj,
1980). Research has also revealed that the acquisition of concrete operational
thinking coincided with a decrease in in-group prejudice, and that conserva-
tion preceded a reduction in prejudice (e.g., Doyle & Aboud, 1995; Doyle et
al., 1988).
On the other hand, however, whereas ST’s account might be taken to
indicate that ethnic prejudice ceases to be a problem during the elementary
school years, or is substantially ameliorated, as a result of children’s increas-
ing cognitive abilities, research suggests that this is not the case. For example,
in studies that have reported an association between conservation and preju-
dice, up to 50% of the children who could conserve still displayed ethnic
prejudice (e.g.., Doyle & Aboud, 1995).
Further, the actual nature of the pattern of children’s responses that ST
seeks to explain is not unambiguous. Although there are certainly a number
of studies that have reported an unambiguous decrease in preference/
prejudice after 7 years of age, as sociocognitive theory would predict (e.g.,
Aboud & Mitchell, 1977; George & Hoppe, 1979; Vaughan, 1964; Williams,
Best, & Boswell, 1975), other studies have reported not only that in-group
preference remained at the same level from 7 to 12 years of age (e.g., Asher &
Allen, 1969; Banks & Rompf, 1973; Davey, 1983; Milner, 1973; Teplin, 1976;
Weiland & Coughlin, 1979) but also that in-group preference actually
increased during these years (e.g., Bartel, Bartel, & Grill, 1973; Hraba &
Grant, 1970; Rice, Ruiz, & Padilla, 1974; Vaughan & Thompson, 1961).
Perhaps of critical importance, however, is the claim noted above that the
224 Social identity and ethnic prejudice
main body of findings that ST seeks to explain may actually consist of
children’s in-group preferences rather than their out-group dislike or prejudice
(e.g., Brand et al., 1974; Katz, 1976; D. Nesdale, 2001a; Proshansky, 1966;
Stephan & Rosenfield, 1979). The basis of this view is that the main methods
commonly used to assess children’s ethnic prejudice, certainly up to the
1990s, rested on children’s choice of ethnically differentiated dolls, photos, or
drawings, or the attribution of traits to such stimuli, and hence might only
index their level of in-group preference (see review by D. Nesdale, 2001a).
Consistent with this are findings indicating a lack of correspondence between
children’s ethnic preferences and their choice of friends and playmates (e.g.,
Fishbein & Imai, 1993; Jansen & Gallagher, 1966), and that children who are
given the opportunity to rate minority group children on bipolar scales (i.e.,
like–dislike) tend to express greater liking for the in-group versus the out-
group, rather than liking for the in-group and dislike for the out-group (e.g.,
Aboud & Mitchell, 1977; D. Nesdale, 1999b). Indeed, contrary to Aboud’s
position, a number of writers have argued that real ethnic prejudice, signified
by, for example, racial cleavage, epithets, or tension, when it does emerge in
children, does not actually appear until children are well into mid childhood,
around 9 or 10 years of age (e.g., Goodman, 1952; Katz, 1976; Milner, 1996;
Proshansky, 1966).
Finally, ST offers a developmental account that is largely indifferent to
the social context and motivational considerations. For example, it is unlikely
that the initiation of prejudice in children is governed simply by the child’s
affective-perceptual processes associated with fear of the unknown (and
attachment to the familiar). On the contrary, some but not all physical differ-
ences are associated with prejudice in both children and adults, the physical
differences to which young children respond are also those of racial signifi-
cance to adults (Katz, Sohn, & Zalk, 1975), and strong prejudices (e.g.,
towards particular national groups, religions) can occur even in the absence
of physical differences (Tajfel, Jahoda, Nemeth, Rim, & Johnson, 1972).
Together, these points emphasise the fact that the differentiation of racial
cues by young children is not determined solely by their perceptual distinct-
iveness based on unfamiliarity. Rather, the cues to which even young children
respond have a distinctiveness that is socially determined, particularly by the
labels and evaluative statements applied to groups by peers and adults (Katz,
1976; Vaughan, 1987).
In sum, although the preceding review is necessarily abbreviated, it is clear
that markedly different theoretical constructions have been built upon the
available body of research findings. Although each approach accounts for
some of the findings, none provides a comprehensive explanation of the
development of children’s prejudice, encompassing within it the full range of
findings that have been revealed to date.
However, the review has served to identify a number of central issues,
which must be taken into account in any comprehensive explanation of the
development of children’s prejudice. First, the theory must account for
 Nesdale 225
the appearance of out-group dislike and hatred versus mere in-group prefer-
ence. Second, the theory must account for the fact that children’s attitudes are
not necessarily mere reflections of those of their parents or siblings. Third, it
must take into account the fact of children’s developing perceptual, cognitive,
and linguistic abilities. Fourth, the theory must give due recognition to social-
motivational issues. Fifth, the theory must account for the fact that some
children develop ethnic prejudice whereas others do not.
Social identity development theory (SIDT) has been devised as a response
to these central issues. Rather than drawing upon individual differences
between children, their supposed tendency to mimic significant others, or
their perceptual-cognitive acquisitions, SIDT pays particular regard to
children’s social motivations and their growing awareness of the nature of
the social world in which they live. On this basis, some of SIDT’s core prop-
ositions are greatly influenced by social identity theory (Tajfel & Turner,
1979).

SOCIAL IDENTITY THEORY AND ETHNIC PREJUDICE

An approach that places considerable emphasis on social-motivational con-

siderations and awareness of social structure in accounting for ethnic preju-
dice is provided by social identity theory (SIT; Tajfel & Turner, 1979) and
its recent elaboration, self-categorisation theory (SCT; Turner et al., 1987).
Briefly, according to SIT, prejudice and discrimination towards members of
ethnic out-groups ultimately derives from the desire of individuals to identify
with social groups that are considered to be positively distinctive or compara-
tively superior to other groups, in order to enhance their own self-esteem. The
consequences of group identification are that in-group members are per-
ceived to be similar and to possess positive qualities and hence are subject to
positive bias. In contrast, out-group members are perceived to be different
and to possess less favourable qualities and hence may attract prejudice and
discrimination. Numerous studies of adults and adolescents have now pro-
vided broad support for SIT (see reviews by Brewer, 1979; R. Brown, 1995;
Hogg & Abrams, 1988; Mullen, Brown, & Smith, 1991), especially in research
using the minimal group paradigm (Tajfel, Billig, Bundy, & Flament, 1971).
However, although SIT is probably the most widely endorsed social-
psychological account of ethnic prejudice in adults at the present time, the
theory is virtually mute on the issue of the development of prejudice in
children. Nevertheless, according to several researchers, there are good
grounds for supposing that SIT might provide the basis of an explanation of
ethnic prejudice in both children and adults (e.g., Davey 1983; Milner, 1996;
D. Nesdale, 1999a, 1999b; D. Nesdale & Flesser, 2001; Vaughan, 1988). For
example, consistent with the basic assumptions of the theory are anecdotal
reports that children from as young as 3 years of age have a developing aware-
ness of which groups in a community are better off and more highly regarded
226 Social identity and ethnic prejudice
than others, and that children make comparisons between their standing as
a member of one ethnic group versus other ethnic groups (e.g., Davey, 1983;
Goodman, 1946; Milner, 1996; Radke & Trager, 1950; Vaughan, 1987). In
addition, the findings of ethnic preference studies are consistent with SIT’s
assumption that people, including children, seek the positive distinctiveness
conferred by membership of higher- rather than lower-status groups in order
to enhance their self-esteem. Thus, whereas dominant-group children rarely
misidentify their ethnic group, members of low-status minority groups
(e.g., black, native, and Hispanic Americans) frequently misidentify with the
dominant (e.g., white American) cultural group (e.g., Asher & Allen, 1969;
Greenwald & Oppenheim, 1968; Hunsberger, 1978; Morland, 1966; Teplin,
1976).
However, while there are some grounds for concluding that the behaviour
of even young children accords with the basic assumptions of SIT, the theory
still has little to say concerning the unique circumstances of growing children.
Issues such as the nature of the instigation of the developmental process in
young children, whether or not there are age-related changes in the acquisi-
tion process, whether these changes are influenced by other cognitive and
linguistic acquisitions, and so on, are not addressed by SIT.
The remainder of this chapter sketches out a model that is designed
to provide a more complete account of the development of prejudice in
children. The model draws upon SIT for its core assumptions, but recognises
that there are changes in children’s intergroup behaviour that are linked to
increasing age.

SOCIAL IDENTITY DEVELOPMENT THEORY OF

CHILDREN’S ETHNIC PREJUDICE

Social identity development theory (SIDT) proposes that children who

display ethnic prejudice pass through four sequential development phases
(undifferentiated, ethnic awareness, ethnic preference, ethnic prejudice).
The phases are differentiated in terms of the behaviours that characterise
them, and the events that precipitate changes from one phase to the
next.

Developmental phases

Phase 1: Undifferentiated
Prior to 2–3 years of age, racial cues are typically not salient to young
children – they respond to objects and people in their environment, initially
in terms of what catches their attention. Increasingly, however, they become
more selective and discriminating and begin to respond differentially to cues
such as gender and age.
 Nesdale 227
Phase 2: Ethnic awareness
Ethnic awareness begins to emerge at around 3 years of age, particularly
among those children who reside in multiracial societies. A number of studies
have confirmed that children can accurately identify and distinguish between
skin colour hues at this age (e.g., K. B. Clark & Clark, 1939; Goodman, 1946;
Stevenson & Stevenson, 1960). As Katz (1976) has emphasised, it is likely
that awareness begins following an adult’s identification/labelling of an out-
group member (e.g., “yes, that person has black skin – he is an Aboriginal/
Afro-American”). It is the perception of such differences, particularly when
accompanied by a verbal label, that is likely to facilitate social categorisation
based on skin colour. It is important to note, however, that young children do
not appear to construct social categories on an idiosyncratic basis (e.g., “yes,
that person has blue shorts/a big nose”). Children enter an environment in
which the key social categories are already specified and the nature of inter-
group relations is established. Accordingly, the social categories that children
are likely to emphasise are not simply those that are strange and unfamiliar
(cf. Aboud, 1988) – they will be those that have social significance in the
community (e.g., Katz, 1976; Vaughan, 1987).
Children’s awareness of these categories will be sharpened by any negative
evaluations communicated by adults, verbally or nonverbally (Milner, 1983).
In addition, it is possible that, in relation to the white and black social
categories, the evaluative associations will be further enhanced by the positive
and negative associations with the colours white and black (e.g., Renninger &
Williams, 1966; Williams & Roberson, 1967) and light and dark (e.g.,
G. Brown & Johnson, 1971; Katz, 1973), respectively. However, while ethnic
awareness may begin to emerge at around 3 years of age, the further refine-
ment, elaboration, and clarification of the child’s concept of a racial/ethnic
group continues over many years, perhaps even up to 10 to 11 years of age,
and appears to comprise a number of age-related phases (see Vaughan, 1963).
A crucially important and early achievement in this sequence concerns the
child’s ethnic self-identification – the realisation that he or she is a member of
a particular group. The evidence suggests that self-identification begins to
occur soon after children become aware of ethnic or racial categories. Accur-
ate ethnic self-identification has been reported in dominant group children
as young as 3 years of age (Marsh, 1970) and in virtually all dominant group
children in multiracial communities by 6 to 7 years of age (see Aboud, 1988,
for a review). It remains unclear whether awareness of a child’s own ethnic
identity precedes or follows his/her awareness of another person’s ethnicity,
although there are reasons to suppose that it is the former (see below). How-
ever, the particular significance of this achievement is that it ushers in the
next phase in the sequence, which overlaps the child’s ongoing development
of ethnic awareness.
228 Social identity and ethnic prejudice
Phase 3: Ethnic preference
Self-identification as a member of the dominant social group comprises a
crucially important piece of a child’s identity jigsaw – in this case, it is a
central piece (together with his/her gender) of the child’s developing social (as
compared with personal) identity (Turner et al., 1987). The child learns that
he or she belongs to, or is a member of, a particular ethnic group.
The major effect of this new understanding is an early focusing on the
in-group rather than the out-group, on similarity rather than difference, on
relative superiority rather than inferiority. That is, consistent with SIT (Tajfel
& Turner, 1979), and ST (Turner et al., 1987), the effect of being categorised
into a group is that children begin an ongoing focus on, and preference for,
their in-group(s). Indeed, research by Katz and her colleagues (e.g., Katz,
1973; Katz & Seavey, 1973; Katz et al., 1975) has revealed that white children
as young as 3 years of age more readily distinguished faces of their own
versus another ethnic group. That is, the effect of ethnic self-identification
and the application of group labels was that out-group faces were actually
made less differentiable and/or accessible.
However, the critically important point to be emphasised here is that
SIDT differs from both ST (Aboud, 1988) and SIT (Tajfel & Turner, 1979) in
terms of the impact of ethnic self-categorisation on the child. Whereas both
ST and SIT assume that ethnic self-caregorisation is sufficient to instigate
both in-group favouring and out-group prejudice responses (Rubin &
Hewstone, 1998), SIDT argues that ethnic self-categorisation mainly activates
a focus on, and accompanying preference for, the in-group.
With the passage of time, children are exposed to and extract information
that is consistent with, and enhances the positive distinctiveness of, their
group. Inevitably, at the same time as children acquire positively discriminat-
ing information about the in-group, they begin to hear and acquire informa-
tion that is (typically) less positive/more negative about comparison ethnic
out-groups. Note that while this information may be retained and may form
the basis of an eventual out-group stereotype, it does not appear to be the
focal concern of children at this age. As noted earlier, ethnic stereotypes do
not appear to consolidate until 8 or 9 years of age. Nevertheless, negative out-
group information may still contribute to their sense of relative in-group
superiority and self-esteem (Milner, 1996).
In short, rather than instigating out-group prejudice (cf. Aboud, 1988;
Tajfel & Turner, 1979), the effect of ethnic self-identification is to instigate an
in-group focus and bias. Consistent with this view is the array of findings that
have emerged in ethnic preference studies (see Nesdale, 2001a, for a review).
For example, when given a forced or restricted choice between in-group and
out-group stimulus figures (e.g., dolls, pictures, drawings), dominant group
children almost invariably indicate a preference for the in-group figure. Simi-
larly, if required to make a choice between assigning positive versus negative
attributes to in-group versus out-group figures, the in-group figure will be
 Nesdale 229
awarded the more positive attributes. While there is an obvious ambiguity
in this data, the present position is that these findings reveal children’s
preference for the in-group rather than dislike or rejection for the out-group.
In addition, research indicates that friendship and playmate preferences
are unrelated to ethnic preferences or out-group stereotype responses (e.g.,
Fishbein & Imai, 1993; Hraba & Grant, 1970), that there is no correlation
between out-group stereotypes and out-group bullying (e.g., Boulton, 1995),
that the negativity of the out-group stereotype drops when the response is
open-ended versus forced choice (e.g., Lerner & Buehrig, 1975; Lerner &
Schroeder, 1971), and that young children rarely give rejection of the out-
group stimulus figure as a reason for their choice of the in-group stimulus
figure (e.g., Zinser, Rich, & Bailey, 1981).
Further, as noted previously, in the few studies in which young children
have given independent responses to in-group and out-group stimulus figures,
and the responses have been on a like–dislike bipolar scale, the children have
almost invariably used only the liking half of the scale. That is, the out-group
stimulus figure has been rated as relatively less likeable than the in-group
stimulus figure, not as disliked (e.g., Aboud & Mitchell, 1977; Genesee,
Tucker, & Lambert, 1978; D. Nesdale, 1999b). Also of importance is the
evidence that ethnicity is typically not an especially salient social category to
young children, and is certainly not as salient as the gender category – friend-
ship and playmate preferences are typically determined by gender, at least up
to 10 or 11 years of age (e.g., Fishbein & Imai, 1993; Helgerson, 1943).
Interestingly, in contrast to ethnicity, research has revealed that children as
young as 4 or 5 years of age are quite prepared to reveal a strong dislike
towards opposite versus same gender stimulus persons – here, the stimulus
figures are unambiguously rated in the “disliked” half of a bipolar scale
(e.g., Yee & Brown, 1994).
In sum, according to the present analysis, ethnic self-identification is a
process that tends to occur in all children, sooner or later. It facilitates and
reflects a growing understanding of the social structure in the community, the
standing of the different groups, and their interrelationships, and the lan-
guage used to describe other group members. For dominant group children it
prompts a focus on, and preference for, the ethnic in-group. (In contrast,
minority group children often reject their in-group in favour of the culturally
dominant out-group.) However, if ethnic preference is merely that (i.e., pref-
erence not prejudice), the question remains as to how ethnic preference turns
into a negative attitude or prejudice.

Phase 4: Ethnic prejudice

Contrary to Aboud’s (1988) claim that ethnic prejudice diminishes in
children from 7 years of age onwards as a result of cognitive acquisitions,
SIDT contends that it is precisely in this period that prejudice actually
crystallises and emerges in those children who come to hold such attitudes. That
230 Social identity and ethnic prejudice
is, according to SIDT, prejudice does not emerge in all children as a matter of
course (cf. Aboud, 1988).
In essence, according to SIDT, prejudice entails an active process of change
from a state of mere ethnic preference. It requires shifts in focus in each of the
child’s perceptual, affective, cognitive, and behavioural domains. Rather than
being focused on the in-group and its positive differentiation from the out-
group, prejudice implies at least an equal focus on in-group and out-group,
if not an obsessive focusing on the out-group. Instead of liking an out-
group member less than an in-group member, prejudice means that out-group
members are disliked or hated. Rather than knowing and being able to repro-
duce (negative) “facts” about ethnic minorities, a prejudiced person holds
them as his/her own. Finally, instead of engaging in interethnic play and
friendship, prejudice means derogating and discriminating against minority
group members whenever the occasion arises.
Clearly, the transition from having a preference for the in-group to feeling
prejudice towards an ethnic minority group is not inconsiderable. According
to SIDT, the transition from ethnic preference to prejudice depends on
several elements.

Acquisition of ethnic constancy

An essential prerequisite of ethnic prejudice appears to be the acquisition of
the concept of ethnic constancy – the understanding that ethnic group
membership is immutable and, like gender, does not change with age (Katz,
1976; Semaj, 1980). As part of children’s developing and elaborating concept
of ethnic group noted earlier, they learn the significant ethnic cues that differ-
entiate groups and they can correctly label positive and negative instances.
Moreover, children learn that ethnic cues, unlike size, are resilient to changes
in age (e.g., Katz, 1976).
Given that ethnic constancy concerns the ability to appreciate the immut-
ability of a quality or construct despite contextual changes, it follows that
ethnic constancy is likely to relate to the acquisition of concrete operational
thinking that occurs in most children at around 7 years of age (e.g., Flavell,
1963). Consistent with this expectation, Semaj (1980) has reported, in a study
with 4- to 11-year-old black children, that conservation of mass and weight
preceded ethnic constancy, which increased with age. The importance of
ethnic constancy is that when it is acquired, the perceptual and cognitive
components of ethnic attitudes are now bought into functional inter-
relationship – minority out-groups (as well as the in-group) now have a
substance and longevity to which negative (or positive) attitudes may be
attached (e.g., Semaj, 1980).
 Nesdale 231
Acquisition of social-cognitive skills
Other things being equal, whether or not children who have ethnic constancy
develop and express prejudice towards minority group members is likely to
be influenced by whether they acquire several important social-cognitive
abilities. These include the ability to decentre and take the perspective of a
minority group child, the ability to empathise and experience the feelings
of such children, and the ability to engage in higher-level moral reasoning
(Feffer & Gourevitch, 1960; Kohlberg, 1976; Selman & Byrne, 1974).
There are good grounds for supposing that these abilities would impact
upon whether children develop ethnic prejudice. For example, it might be
anticipated that dominant group children who are able to decentre and per-
ceive the social environment from the perspective of a minority group child,
and to share the feelings of the latter, would be less likely to develop prejudice
towards members of that out-group. Similarly, it might be supposed that
children whose moral judgments go beyond the external consequences of
actions and the individual’s need satisfactions to focus on right as defined
by social authority or universal principles, would also be less likely to develop
ethnic prejudice. However, although there is some indirect evidence consistent
with these speculations (e.g., A. Clark, Hocevar, & Dembo, 1980; Madge,
1976), the impact of children’s developing social-cognitive abilities on their
ethnic prejudice remains to be assessed directly.

Social identity processes

While the acquisition of sociocognitive abilities such as decentration,
empathy, and moral reasoning are certainly contrary to the development of
prejudice, SIDT proposes that the principal determinant of children’s preju-
dice is a social process. That is, instead of merely preferring the in-group,
dominant group children change to disliking minority out-groups when they
adopt the negative ethnic attitudes that prevail in their social group.
As noted earlier, however, children do not simply ape the ethnic attitudes
and behaviours of those around them. Rather, they begin actively disliking
ethnic minorities when they adopt, as their own, the negative out-group atti-
tudes that prevail among those people whom they value and with whom they
identify, in their social environment. That is, children adopt a particular atti-
tude because it fits with their view of themselves as belonging to a social
group with a particular set of attitudes, beliefs, and behaviours, and they
derive positive distinctiveness from that group membership (Milner, 1996).
It is important to note that the change from preference to prejudice is
unlikely to be immediate and would typically represent the culmination of a
period of exposure to the dislike or hatred felt by significant others towards
minority group members, the negative “facts” (i.e., stereotypic beliefs) that
are espoused in relation to minority group members, and the observation
of discriminatory behaviours directed at them. According to a number of
232 Social identity and ethnic prejudice
writers, children are capable of reproducing these “second-hand” learnings
during the ethnic preference phase but, at that stage, they are not the
children’s own and typically do not impact upon their play and friendship
preferences – at this age, children “do not walk the talk” (e.g., Chyatte,
Schaefer, & Spiaggia, 1951; Goodman, 1946; D. Nesdale, 2001b; Proshansky,
1966; Radke et al., 1949).
SIDT proposes that three factors facilitate the change from ethnic prefer-
ence to prejudice. First, the probability of children adopting an ethnic preju-
dice as their own will increase to the extent that children identify with a
(dominant) group in which prejudice is widely shared by the group members.
As consensus increases, prejudice will become increasingly normative, with
concomitant expectations that it will be adopted by all in-group members.
Second, the tendency for children to develop ethnic prejudice will increase
as tension and threat increases between members of the dominant and ethnic
minority group(s) (Brown, 1995). Under these circumstances, in-group iden-
tification and bias will increase, giving rise to an increasing tendency towards
out-group rejection and dislike.
Third, the tendency towards ethnic prejudice in children will be greatest
when there is a state of conflict between members of the two groups.
Although evidence consistent with the preceding three predictions has been
reported in relation to adults (e.g., Branscombe & Wann, 1994; Hamm, 1993;
Levin & McDevitts, 1993; Long, Spears, & Manstead, 1994; Quillian, 1995;
Stephan, Ybarra, Martinez, Schwarzwald, & Tur-Kaspa, 1998), the impact
of these factors on children’s ethnic prejudice remains to be assessed.

Implications of SIDT
Several implications follow from SIDT. For example, whereas positive correl-
ations between the ethnic attitudes of children and their parents might nor-
mally be expected, this need not necessarily be so. Indeed, Pushkin (in Davey,
1983) found a closer relationship between children’s attitudes and those of
people in the neighbourhood, than between the children’s attitudes and those
of their parents. Further, several studies have reported that the children of
black activist parents displayed greater pro-white preference than the children
of less active parents (Branch & Newcombe, 1980; Floyd, in Williams &
Morland, 1976). This pattern of findings is clearly supportive of the view that
the particular ethnic attitudes adopted by children reflect their own interests
and perceptions rather than those who might either be expected to or seek to
influence them.
A further implication is that the factors described above, which are pro-
posed to enhance children’s social identification processes, are likely to over-
whelm the sociocognitive acquisitions of role-taking and moral reasoning, at
least in relation to particular disliked ethnic out-groups. Consistent with this
are several studies reporting that 7-year-old children were accurate at taking
the role of another ethnic person, provided that the out-group person was a
 Nesdale 233
member of a liked group. If that person was disliked, then the role-taking was
inaccurate (Aboud & Mitchell, 1977; Middleton, Tajfel & Johnson, 1970).
According to Aboud and Mitchell, such findings reflect the impact of nega-
tive attitudes on the utilisation of perceptual cues, whereas Middleton et al.
considered that a negative attitude prevents a child from circumventing his
egocentric tendencies. Either way, the findings endorse the greater influence
exerted by social and motivational factors over social-cognitive factors.
Another implication is that since SIDT is primarily founded upon social-
motivational rather than perceptual-cognitive considerations, it would not
predict that the appearance of ethnic prejudice in children would be linked to
specific ages (cf. Aboud, 1988). According to SIDT, although prejudice is
unlikely to occur in children younger than 6 or 7 years of age because their
cognitive abilities would not have achieved the requisite level of development,
its emergence thereafter would be dependent on their unique social situation.
At any time, children’s attitudes towards members of ethnic out-groups
might increase, decrease, or remain the same, depending on their prevailing
social group identification.
A final implication of the present model is that children (and adults) may
never display ethnic prejudice because they choose not to identify with a
social group that has a negative attitude towards an ethnic minority group(s).
Interestingly, this does not necessarily imply that such children would not
continue to prefer their in-group(s) over other out-groups for, at one level,
identification means preference, not prejudice (see also R. Brown, 1995).

Recent research support for SIDT

Although SIDT provides a good account of the results of the extant ethnic
preference and trait attribution studies, it needs to be remembered that these
paradigms have distinct limitations, if not outright flaws. In particular, as
noted above, these paradigms typically necessitate a choice being made
between photos or dolls representing members of different ethnic groups. In
addition, the transparency of the intent of these procedures enhances the
likelihood of children giving socially desirable responses, even in the more
contemporary versions of the technique (e.g., Black-Gutman & Hickson,
1996: Boulton, 1995; Doyle et al., 1988), particularly as children increase in
age. On this basis, more recent research has typically utilised the minimal
group paradigm in which the participants are randomly assigned member-
ship in one of several groups (Tajfel et al., 1971). The advantage of this
technique in terms of research on children’s intergroup attitudes is that their
responses can be examined in relation to the members of their in-group, as
well as to members of out-groups, rather than simply to an individual.
Of particular importance to the present discussion is the fact that recent
research using this paradigm has focused on identifying when children
reveal in-group identification, what factors promote it, if and when
children are responsive to status differences between groups, what effect
234 Social identity and ethnic prejudice
group membership has on children, and what factors contribute to children
changing from mere ethnic preference to out-group prejudice.
For example, consistent with SIDT, Bigler and her colleagues revealed that
the random assignment of 6- to 11-year-old children to groups (e.g., “red”
group, “green” group) in a minimal group paradigm prompted in-group
favouritism in the children in both groups, regardless of age and gender.
Compared with control group children, those assigned to colour groups did
not want to change groups, rated their own group as most likely to win a
series of three contests, and chose more members of the in-group versus the
out-group to participate in a field trip (Bigler, 1995; Bigler, Jones, & Lobliner,
1997). Bigler et al. also reported that children with higher levels of self-esteem
showed higher levels of intergroup stereotyping.
Further, while these researchers explicitly sought to de-emphasise social
comparisons and competitiveness between the groups, other research has
indicated that when the latter are emphasised, in-group favouritism increases
(Vaughan, Tajfel, & Williams, 1981; Yee & Brown, 1992). For example,
Vaughan et al. emphasised intergroup comparison in another minimal group
study with 7- and 11-year-old children who had also been randomly assigned
to groups. They found that the children allocated rewards to in-group and
out-group participants so as to maximise the difference between the groups,
in favour of the in-group.
Importantly, in their study D. Nesdale and Flesser (2001) sought to exam-
ine the effect of status differences that exist between members of dominant
and minority groups. They randomly assigned 5- and 8-year-old children to
teams that supposedly varied in drawing ability (i.e., high versus low), in
order to manipulate social status. Their results indicated that even the 5-year-
old children were sensitive to the status of their social group, that they liked
their group more, and that they saw themselves as being more similar to the
in-group members, when their group had high versus low status. In addition,
low- versus high-status group members sought to change their group mem-
bership and, like adults, their liking and similarity ratings of in-group and
out-group members were influenced by whether they could change groups.
However, although the preceding findings are clearly consistent with
SIDT’s predictions, it is important to recognise that these studies focused on
children’s responses to contrived or invented groups rather than to particular
ethnic out-groups. Accordingly, to provide a further test of SIDT, D. Nesdale,
Durkin, Maass, and Griffiths (2001b) examined 5-, 7-, and 9-year-old child-
ren’s attitudes to members of particular ethnic groups that differed in status.
Intergroup status was manipulated as in the research by D. Nesdale and
Flesser (2001), and each Anglo-Australian participant was allocated to a
same-sex plus same-ethnicity (i.e., Anglo-Australian) team whereas the out-
group competitor team was revealed (via photographs) to be comprised of
Anglo-Australian children or Pacific Islander children. Again, consistent
with SIDT, the results confirmed that children as young as 5 years of age
are sensitive to status differences between groups and that they wish to be
 Nesdale 235
members of the high- versus low-status group. In addition, the findings also
confirmed the importance of self-categorisation – that children liked the
members of their in-group more than out-group members. Most importantly,
the findings revealed that this effect was enhanced when the out-group
was comprised of members who shared a different ethnic background. The
ethnicity difference apparently sharpens and accentuates the category differ-
ence, with the effect that different-ethnicity out-group members are liked less
than same-ethnicity out-group members. At the same time, it is important to
note that the children did not actually reveal dislike towards the different-
ethnicity out-group members. Instead, the latter group was attributed less
liking than the same-ethnicity out-group. That is, the children’s ratings of the
out-group, regardless of its ethnicity, were never lower than the midpoint of
the bipolar scale. According to SIDT, these findings would confirm that the
children were actually in the ethnic preference phase, with their major focus
being the interests of the in-group, not directing prejudice towards an ethnic
out-group.
A more recent study by D. Nesdale, Durkin, Griffiths, and Maass (2001a)
has further extended these findings. Drawing upon SIDT, the reasoning
behind this study was that if it is the case that children are primarily con-
cerned with their in-group, they will not tend to respond to out-group eth-
nicity unless it happens to be coincident with the boundary between the two
groups (i.e., the in-group’s ethnicity differs from that of the out-group).
Under the latter circumstances, differential ethnic preference would be
expected to be displayed as a result of the category difference. The implica-
tion here is that, whereas children in the ethnic preference phase will like a
different ethnicity out-group less than a same ethnicity out-group, a different
ethnicity in-group (i.e., one that contains members of different ethnicity to
the target child) will be liked no less than a same ethnicity in-group.
To test this reasoning, 5-,7-, and 9-year-old Anglo-Australian children par-
ticipated in a minimal group study in which they were assigned to a high-
status in-group that included other Anglo-Australian or Pacific Islander
children. As in the previous study, the children also expected to compete with
an out-group that contained Anglo-Australian or Pacific Islander children.
Consistent with SIDT, and with D. Nesdale et al. (2001b), children in a
same-ethnicity (i.e., Anglo-Australian) in-group indicated greater liking for
the in-group over the out-group, but more so when the latter was comprised
of members of the same (i.e., Anglo-Australian) versus a different ethnicity
(i.e., Pacific Islander) out-group. However, the inclusion of a different-
ethnicity (i.e. Pacific Islander) in-group yielded a critically important and dif-
ferent pattern of findings. First, as predicted, having in-group members
whose ethnicity differed from one’s own had no impact on the children’s
liking for the in-group. That is, the children rated the same and different in-
groups as being equally likeable, and more likeable than the out-group.
Clearly, these findings indicate that, depending on the context, ethnicity sim-
ply does not matter to children, at least up to 9 years of age: If different
236 Social identity and ethnic prejudice
ethnicity children happen to be members of the in-group, then the in-group is
liked no less than one that is made up of same-ethnicity members. As speci-
fied by SIDT (and with SIT, Tajfel & Turner, 1979), it is the categorisation
that matters, not the ethnicity per se.
Second, and compared with the same-ethnicity in-group results, the
analysis revealed that when the in-group was made up of different ethnicity
members, the children liked the same-ethnicity out-group less than the
different-ethnicity out-group. In short, the children’s standard of comparison
appeared to have changed. Rather than seeing different-ethnicity out-group
members as markedly different, as occurred when the in-group was com-
prised of same-ethnicity members, in the case of a different-ethnicity in-
group, the members of a different-ethnicity out-group were perceived as less
different than a same-ethnicity out-group. Again, consistent with SIDT, this
finding emphasises that in the absence of threat and conflict, young children
up to 9 years of age are simply not repositories of ethnic dislike and prejudice
(cf. Aboud, 1988).
Although the results of the preceding studies have emphasised the impact
of ethnic self-categorisation and intergroup status differences on children,
even those as young as 5 years of age, and provides considerable support for
SIDT, additional light is shed on these processes by two further studies. Con-
trary to the preceding research, the latter studies have not employed the
minimal group paradigm. Instead, these studies sought to examine the effect
of different sorts of information on children’s attitudes towards particular
members of their own and other ethnic groups, using a paradigm that min-
imised the transparency of the goal of the research by reducing the salience
of ethnicity and the focus on prejudice. This technique (e.g., A.R. Nesdale &
McLaughlin, 1987) involves children reading (or being read) a short story
involving two characters, one being of the same, and the other of different,
ethnicity to the subject. The story is thematic (e.g., “a day at the zoo”) and
each character reveals a particular set of traits and behaviours as the story
unfolds. Thus, the salience of ethnicity, as well as the focus on prejudice, are
both de-emphasised and the task is made more familiar and realistic because
of the array of information presented, as well as the other issues that are
addressed (e.g., what did the characters wear, what did they do, etc.), in
addition to how much the children liked the story characters.
The first study (Nesdale, 1999b) was designed to examine the interplay
between children’s cognitive abilities and social motivations, as the children
increased in age from 8 to 10 to 12 years. The Anglo-Australian children
listened to a story about an in-group Anglo-Australian boy and an out-group
Vietnamese boy, each of whom displayed equal numbers of the relevant
ethnic stereotype-consistent and stereotype-inconsistent traits. Each story
character also displayed a positive and a negative behaviour. The results
indicated that although the children’s responses changed with increasing age,
reflecting their increasing cognitive abilities, their responses were not merely
dependent on their expanding abilities to differentiate individuals. Instead,
 Nesdale 237
consistent with SIDT, the children’s attention was not shared equally between
the in-group and out-group characters. Of particular significance to the
children as they increased in age was the stereotype inconsistency of
the in-group story character.
Thus, the results revealed that, as they increased in age, the children
remembered more of the in-group versus the out-group story character’s
stereotype-inconsistent traits and that they increasingly disliked the in-group
story character. In addition, the in-group and out-group story characters’
negative behaviours were attributed to internal and external causes, respect-
ively, whereas their positive behaviours were attributed to external and
internal causes, respectively. In short, rather than diminishing their interest
in, and preference for, their in-group as they increased in age, the children’s
responses were increasingly motivated by their concerns regarding the
in-group character’s worthiness to be a member of their in-group.
What is also noteworthy, however, is that this in-group orientation did not
necessarily lead them to reject, dislike, or express prejudice towards the out-
group member as a matter of course. Indeed, the older children actually
expressed more liking for the out-group than the in-group story character.
This response is certainly consistent with the ethnic preference phase speci-
fied by SIDT and has been reported as the “black sheep” effect in adults
(e.g., Marques, Robalo, & Rocha, 1992; Marques, Yzerbyt, & Leyens, 1988).
The second study (D. Nesdale & Brown, 2001) utilised the same paradigm
to develop this line of research further. Since previous research has shown
that children’s in-group versus out-group stereotypes are almost invariably
positive (A. R. Nesdale, 1987), one possibility was that converting half the in-
group’s stereotypic traits to counterstereotypic traits in the D. Nesdale
(1999b) study might have given rise to an in-group character who was espe-
cially unattractive, particularly in comparison with the out-group character.
Accordingly, in this study, 6-, 9-, and 12-year-old Anglo-Australian children
listened to a story about an Anglo-Australian and a Chinese boy, each of
whom displayed two positive and two negative traits that pilot research had
revealed were matched in positivity and negativity, respectively.
In addition, in order to assess whether SIDT’s distinction between ethnic
preference and ethnic prejudice could be demonstrated empirically, the study
also sought to manipulate the relationship between the two story characters.
Thus, for half the children, the story characters were revealed to be the best
of friends who shared lots of activities. In contrast, for the remaining chil-
dren, the story characters were described as being bad friends whose relation-
ship had been characterised by lots of disputes. Consistent with D. Nesdale
(1999b), it was assumed that ethnic preference would be instigated when the
story characters had a positive relationship, but that this might turn to dis-
like/prejudice when the relationship between the out-group character and
favoured in-group character was one of enmity (i.e., due to their shared group
membership, the children would join with the in-group character in disliking
the out-group character).
238 Social identity and ethnic prejudice
However, the results revealed that, regardless of the story characters’ rela-
tionship, the children gave ethnic preference responses. As they increased in
age, the children remembered more of the in-group character’s negative
versus positive traits, saw themselves as increasingly dissimilar to him,
and they liked him less. In contrast, with increasing age, the children
remembered more of the out-group character’s positive versus negative traits,
saw themselves as increasingly similar to him, and liked him more.
Although the preceding findings are consistent with those predicted by
SIDT for children in the ethnic preference phase, the negative relationship
manipulation apparently did not instigate ethnic prejudice. That is, it did not
enhance the children’s identification with the in-group character and result in
them, in consort with the in-group story character, feeling threatened by
the enmity from the out-group story character. Rather, the story characters’
relationship was simply not seen as being greatly relevant to the children’s
primary focus on, and preference for, the in-group. Clearly, the implication
from these findings is that a different and stronger manipulation will be
required in order to test SIDT’s proposals concerning the transition of
children from ethnic preference to prejudice.

CONCLUSIONS

To date, three main accounts of children’s ethnic prejudice have been pro-
posed. These accounts include the view that ethnic prejudice in children is a
form of emotional maladjustment arising from a faulty parent–child rela-
tionship; that ethnic prejudice emerges in children as a reflection of the views
of significant others; and that ethnic prejudice is determined by changes in
children’s perceptual-cognitive abilities. Although each approach accounts
for some of the extant findings, the present review suggests that none provides
a comprehensive explanation of the development of children’s ethnic
prejudice.
In particular, the present analysis argues that the critical issues upon which
the preceding approaches fall short concern the need to differentiate in-group
preference from out-group prejudice, the fact that children simply do not as a
matter of course adopt as their own the ethnic attitudes and behaviours of
their parents and peers, the fact that changes in perceptual and cognitive
abilities do not straightforwardly determine ethnic attitudes, and the
importance of children’s need to be members of valued social groups.
In contrast to the extant approaches, social identity development theory
(SIDT) has been designed to take these issues into account. Importantly,
SIDT emphasises the critical significance of social identity processes in the
development of children’s ethnic attitudes and, in so doing, facilitates a
long-overdue shift away from the prevailing emphasis in much social-
developmental research on the predominance of cognitive processes.
As the discussion has revealed, the case for SIDT is compelling. The theory
 Nesdale 239
provides a good fit for the findings that have been revealed via ethnic prefer-
ence and trait attribution studies, as well as the results obtained using other
paradigms. In addition, support for the theory has been strengthened by the
results from newer paradigms that have enabled the intensity of children’s
intergroup attitudes to be assessed. The results of the latter studies have
supported many of the main tenets of SIDT.
In particular, the results have confirmed that children as young as 5 years
of age are sensitive to status differences between groups and that they wish to
be members of high- versus low-status groups. The findings have also con-
firmed the impact of self-categorisation on children. The evidence indicates
that self-categorisation gives rise to a considerable positive orientation to the
in-group. The effect is that children feel themselves to be similar to and they
like the members of their in-group. Indeed, children like different-ethnicity
in-group members as much as same-ethnicity in-group members, and both
are liked more than out-group members. Further, their positive orientation to
the in-group is such that if in-group members are perceived to behave in a
manner inappropriate to the group, they will be seen as “black sheep” and
liked less than an equally negative out-group member.
In contrast to their attitudes towards the in-group, the findings have
revealed that the children’s liking for out-group members may be influenced
by the ethnicity of the out-group. If the in-group and out-group are differen-
tiated by ethnicity, then the ethnicity difference apparently sharpens and
accentuates the category difference, with the effect that different-ethnicity
out-group members are liked less than same-ethnicity out-group members.
However, consistent with SIDT, it is important to note that children reveal
less liking rather than dislike or hatred towards out-group members, even
those of differing ethnicity, in the absence of intergroup threat or conflict.
Although the preceding findings provide good support for SIDT, there are
several important issues that still require resolution. For example, of central
interest is the issue of how young children’s status preferences relate to their
ethnic attitudes. The results of several studies (e.g., D. Nesdale & Flesser,
2001; Yee & Brown, 1992) indicate that children do wish to be members of
high-status groups. However, in the two studies in which status was manipu-
lated, and group membership was realistically displayed via photographs (i.e.,
Yee & Brown, 1992; D. Nesdale et al., 2001b), status had no impact on liking.
Indeed, even in the study in which group status was revealed to have an effect
on in-group liking (i.e., D. Nesdale & Flesser, 2001), status had no effect on
out-group liking, as might be predicted by SIDT. That is, children in the
low-status in-group did not evidence an equal or greater liking for the high-
status out-group; instead, they still preferred their in-group. Although these
findings suggest that children are attracted by group status, but that their
liking is determined by group membership, this issue requires more research
attention.
Another important issue concerns the nature of the transition process
between the ethnic preference and ethnic prejudice phases, as proposed by
240 Social identity and ethnic prejudice
SIDT. To date, most of the research has focused on children in the ethnic
preference phase. The questions that are now of focal concern relate to the
specification of the factors that instigate the transition, as well as the nature
of the changes in children as they move from a state of ethnic preference to
ethnic prejudice. At present, little research has addressed these issues. The
final assessment of the viability of SIDT as an account of the development of
children’s prejudice waits upon the outcome of this research.

REFERENCES

Aboud, F. E. (1984). Social and cognitive bases of ethnic identity constancy. Journal
of Genetic Psychology, 184, 217–230.
Aboud, F. (1988). Children and prejudice. Oxford: Blackwell.
Aboud, F., & Doyle, A. (1996a). Does talk of race foster prejudice or tolerance in
children? Canadian Journal of Behavioural Science, 28, 161–170.
Aboud, F., & Doyle, A. (1996b). Parental and peer influences on children’s racial
attitudes. International Journal of Intercultural Relations, 20, 371–383.
Aboud, F. E., & Mitchell, F. G. (1977). Ethnic role taking: The effects of preference
and self-identification. International Journal of Psychology, 12, 1–17.
Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, R. N. (1950). The
authoritarian pesonality. New York: Harper & Row.
Allport, G. W. (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley.
Asher, S. R., & Allen, V. L. (1969). Racial preference and social comparison processes.
Journal of Social Issues, 25, 157–167.
Banks, W. C., & Rompf, W. J. (1973). Evaluative bias and preference behaviour in
black and white children. Child Development, 44, 776–783.
Bartel, H. W., Bartel, N. R., & Grill, J. J. (1973). A sociometric view of some
integrated open classrooms. Journal of Social Issues, 29, 159–173.
Bigler, R. S. (1995). The role of classification skill in moderating environmental
influences on children’s gender stereotyping: A study of the functional use of
gender in the classroom. Child Development, 66, 1072–1087.
Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorisation and the
formation of intergroup attitudes in children. Child Development, 68, 530–543.
Bigler, R. S., & Liben, L. S. (1993). A cognitive-developmental approach to
racial stereotyping and constructive memory in Euro-American children. Child
Development, 64, 1507–1518.
Bird, C., Monachesi, E. D., & Burdick, H. (1952). Infiltration and the attitudes of
white and Negro parents and children. Journal of Abnormal Social Psychology, 47,
688–689.
Black-Gutman, D., & Hickson, F. (1996). The relationship between racial attitudes
and social-cognitive development in children: An Australian study. Developmental
Psychology, 32, 448–456.
Boulton, M. (1995). Patterns of bully/victim problems in mixed race groups of
children. Social Development, 4, 277–293.
Branch, C. W., & Newcombe, N. (1980). Racial attitude development among young
black children as a function of parental attitudes: A longitudinal and cross-section
study. Child Development, 57, 712–21.
 Nesdale 241
Brand, E. S., Ruiz, R. A., & Padilla, A. M. (1974). Ethnic identification and
preference: A review. Psychological Bulletin, 81, 860–890.
Branscombe, N. R., & Wann, D. L. (1994). Collective self-esteem consequences of
outgroup derogation when a valued social identity is on trial. European Journal of
Social Psychology, 24, 641–657.
Brewer, M. B. (1979). Ingroup bias in the minimal intergroup situation: A cognitive
motivational analysis. Psychological Bulletin, 86, 307–324.
Brown, G., & Johnson, S. P. (1971). The attribution of behavioural connotations to
shaded and white figures by Caucasian children. British Journal of Social and
Clinical Psychology, 10, 306–312.
Brown, R. (1995). Prejudice: Its social psychology. Oxford: Basil Blackwell.
Cameron, J. A., Alvarez, J. M., Ruble, D. N., & Fuligni, A. J. (2001). Children’s lay
theories about ingroups and outgroup: Reconceptualizing research on prejudice.
Personality and Social Psychology Review, 5, 118–128.
Chyatte, C., Schaefer, D. F., & Spiaggia, M. (1951). Prejudice verbalisation among
children. Journal of Educational Psychology, 42, 421–431.
Clark, A., Hocevar, D., & Dembo, M. H. (1980). The role of cognitive development in
childen’s explanations and preferences for skin colour. Developmental Psychology,
16, 332–339.
Clark, K. B., & Clark, M. K. (1939). Segregation as a factor in the racial identification
of Negro pre-school children: A preliminary report. Journal of Experimental
Psychology, 8, 161–163.
Davey, A. (1983). Learning to be prejudiced. London: Edward Arnold.
Doyle, A. B., & Aboud, F. E. (1995). A longitudinal study of white children’s
racial prejudice as a social-cognitive development. Merrill-Palmer Quarterly, 41,
209–228.
Doyle, A. B., Beaudet, J., & Aboud, F. E. (1988). Developmental patterns in the
flexibility of children’s ethnic attitudes. Journal of Cross-Cultural Psychology, 19,
3–18.
Durkin, K. (1995). Developmental social psychology: From infancy to old age. Oxford:
Blackwell.
Feffer, M., & Gourevitch, V. (1960). Cognitive aspects of role-taking in children.
Journal of Personality, 28, 383–396.
Fishbein, H. D., & Imai, S. (1993). Preschoolers select playmates on the basis of
gender and race. Journal of Applied Developmental Psychology, 14, 303–316.
Flavell, J. H. (1963). The developmental psychology of Jean Piaget. New York: Litton
Educational Publishing.
Frenkel-Brunswik, E., & Havel, J. (1953). Prejudice in the interviews of children:
Attitudes toward minority groups. Journal of Genetic Psychology, 82, 91–136.
Genessee, F., Tucker, G. R., & Lambert, W. E. (1978). The development of ethnic
identity and ethnic role-taking skills in children from different school settings.
International Journal of Psychology, 13, 39–57.
George, D. M., & Hoppe, R. A. (1979). Racial identification, preference, and
self-concept. Journal of Cross-Cultural Psychology, 10, 85–100.
Goodman, M. (1946). Evidence concerning the genesis of interracial attitudes.
American Anthropologist, 48, 624–630.
Goodman, M. (1952). Race awareness in young children. Cambridge, MA:
Addison-Wesley.
Greenwald, H. J., & Oppenheim, D. B. (1968). Reported magnitude of
242 Social identity and ethnic prejudice
self-misidentification among negro children – artifact? Journal of Personality and
Social Psychology, 8, 49–52.
Gregor, A. J., & McPherson, D. A. (1966). Racial attitudes among White and Negro
children in a deep-south standard metropolitan area. Journal of Social Psychology,
68, 95–106.
Hamm, M. S. (1993). American skinheads: The criminology and control of hate crime.
Westport, CT: Praeger.
Harris, D., Gough, H., & Martin, W. E. (1950). Children’s ethnic attitudes. II:
Relationships to parental beliefs concerning child training. Child Development, 21,
169–181.
Helgerson, E. (1943). The relative significance of race, sex, and facial expression in
choice of playmate by the preschool child. Journal of Negro Education, 12, 617–622.
Hogg, M. A., & Abrams, D. (1988). Social identifications: A social psychology of
intergroup relations and group processes. London: Routledge.
Horowitz, E. L. (1936). The development of attitude toward the Negro. Archives
of Psychology, 194, 2–48.
Horowitz, E. L., & Horowitz, R. E. (1938). Development of social attitudes in
children. Sociometry, 1, 301–338.
Hraba, J., & Grant, G. (1970). Black is beautiful: A reexamination of racial preference
and identification. Journal of Personality and Social Psychology, 16, 398–402.
Hunsberger, B. (1978). Racial awareness and preference of White and Indian
Canadian children. Canadian Journal of Behavioural Science, 10, 176–179.
Jansen, V. G., & Gallagher, J. J. (1966). The social choices of students in racially
integrated classes for the culturally disadvantaged talented. Exceptional Children,
33, 221–226.
Katz, P. A. (1973). Perception of racial cues in preschool children: A new look.
Developmental Psychology, 8, 295–299.
Katz, P. A. (1976). The acquisition of racial attitudes in children. In P. A. Katz (Ed.),
Towards the elimination of racism (pp. 125–154). New York: Pergamon Press.
Katz, P. A., & Seavey, C. (1973). Labels and children’s perception of faces. Child
Development, 44, 770–775.
Katz, P. A., Sohn, M., & Zalk, S. R. (1975). Perceptual concomitants of racial
attitudes in urban grade-school children. Developmental Psychology, 11, 135–144.
Kohlberg, L. (1976). Moral stages and moralisation: The cognitive-developmental
approach. In T. Lickona (Ed.), Moral development and behaviour. New York: Holt,
Rinehart & Winston.
Lerner, R. M., & Buehrig, C. J. (1975). The development of racial attitudes in young
Black and White children. Journal of Genetic Psychology, 127, 45–54.
Lerner, R. M., & Schroeder, C. (1971). Kindergarten children’s active vocabulary
about body build. Developmental Psychology, 5, 179.
Levin, J., & McDevitts, J. (1993). Hate crimes: The rising tide of bigotry and bloodshed.
New York: Plenum Press.
Livesley, W. J., & Bromley, D. B. (1973). Person perception in childhood and
adolescence. New York: John Wiley & Sons.
Long, K. M., Spears, R., & Manstead, A. S. R. (1994). The influence of personal and
collective self-esteem on strategies of social differentiation. British Journal of Social
Psychology, 33, 313–319.
Madge, N. J. H. (1976). Context and the expressed ethnic preferences of infant school
children. Journal of Child Psychology and Psychiatry, 17, 337–344.
 Nesdale 243
Marques, J. M., Robalo, E. M., & Rocha, S. A. (1992). Ingroup bias and the “black
sheep” effect: Assessing the impact of social identification and perceived variability
on group judgments. European Journal of Social Psychology, 22, 331–352.
Marques, J. M., Yzerbyt, V. Y., & Leyens, J. P. (1988). The “black sheep effect”:
Extremity of judgments towards in-group members as a function of group identifi-
cation. European Journal of Social Psychology, 18, 1–16.
Marsh, A. (1970). Awareness of racial differences in West African and British
children. Race, 11, 289–302.
Middleton, M. R., Tajfel, H., & Johnson, N. B. (1970). Cognitive and affective aspects
of children’s national attitudes. British Journal of Social and Clinical Psychology, 9,
122–134.
Milner, D. (1973). Racial identification and preference in “black” British children.
European Journal of Social Psychology, 3, 281–295.
Milner, D. (1983). Children and race: Ten years on. London: Ward Lock Educational.
Milner, D. (1996). Children and racism: Beyond the value of the dolls . . . In W. Peter
Robinson (Ed.), Social groups and identities: Developing the legacy of Henri Tajfel.
Oxford: Butterworth-Heinemann.
Morland, J. K. (1966). A comparison of race awareness in Northern and Southern
children. American Journal of Orthopsychiatry, 36, 22–31.
Mosher, D. L., & Scodel, A. (1960). Relationships between ethnocentrism in children
and the ethnocentrism and authoritarian rearing practices of their mothers. Child
Development, 31, 369–376.
Mullen, B., Brown, R., & Smith, C. (1991). Ingroup bias as a function of salience,
relevance and status: An integration. European Journal of Social Psychology, 22,
103–122.
Nesdale, A. R. (1987). Ethnic stereotypes and children. Multicultural Australia Papers,
Whole No. 57.
Nesdale, A. R., & McLaughlin, K. (1987). Effects of sex stereotypes on young
children’s memories, predictions and liking. British Journal of Developmental
Psychology, 5, 231–241.
Nesdale, D. (1999a). Social identity and ethnic prejudice in children. In P. Martin &
W. Noble (Eds.), Psychology and society (pp. 92–110). Brisbane: Australian
Academic Press.
Nesdale, D. (1999b). Developmental changes in children’s ethnic preferences and
social cognitions. Journal of Applied Developmental Psychology, 20, 501–519.
Nesdale, D. (2001a). Development of prejudice in children. In M. Augoustinos &
K. Reynolds (Eds.). Understanding prejudice, racism, and social conflict. London:
SAGE Publications.
Nesdale, D. (2001b). Language and the development of children’s ethnic prejudice.
Journal of Language and Social Psychology, 20, 90–110.
Nesdale, D., & Brown, K. (2001). Development of children’s ethnic attitudes.
Unpublished manuscript, Griffith University.
Nesdale, D., & Flesser, D. (2001). Social identity and the development of children’s
group attitudes. Child Development, 72, 506–517.
Nesdale, D., Durkin, K., Griffiths, J., & Maass, A. (2001a). Effects of ingroup and
outgroup ethnicity on children’s attitudes towards the ingroup and outgroup.
Unpublished manuscript, Griffith University.
Nesdale, D., Durkin, K., Maass, A., & Griffiths, J. (2001b). Social identity processes
and children’s ethnic attitude. Unpublished manuscript, Griffith University.
244 Social identity and ethnic prejudice
Peevers, B. H., & Secord, P. F. (1973). Developmental changes in attribution of
descriptive concepts to persons. Journal of Personality and Social Psychology, 27,
120–128.
Porter, J. D. R. (1971). Black child, white child: The development of racial attitudes.
Cambridge, MA: Harvard University Press.
Proshansky, H. M. (1966). The development of intergroup attitudes. In
L. W. Hoffman & M. L. Hoffman (Eds.), Review of child development research
(pp. 311–371). New York: Russell Sage.
Quillian, L. (1995). Prejudice as a response to perceived group threat: Population
composition and anti-immigrant and racial prejudice in Europe. American
Sociological Review, 60, 586–611.
Radke, M. J., & Trager, H. G. (1950). Children’s perceptions of the social roles of
Negroes and whites. Journal of Psychology, 29, 3–33.
Radke, M. J., Trager, H. G., & Davis, H. (1949). Social perceptions and attitudes
of children. Genetical Psychology Monograms, 40, 327–447.
Radke-Yarrow, M., Trager, H., & Miller, J. (1952). The role of parents in the
development of children’s ethnic attitudes. Child Development, 23, 13–53.
Renninger, C. A., & Williams, J. E. (1966). Black–white colour connotations and race
awareness in pre-school children. Perceptual and Motor Skills, 22, 771–785.
Rice, A. S., Ruiz, R. A., & Padilla, A. M. (1974). Person perception, self-identify, and
ethnic group preference in Anglo, Black and Chicano preschool and third grade
children. Journal of Cross-Cultural Psychology, 5, 100–108.
Rosenfield, D., & Stephan, W. G. (1981). Intergroup relations among children. In S. S.
Brehm, S. M. Kassin, & F. X. Gibbons (Eds.), Developmental social psychology
(pp. 271–297). New York: Oxford University Press.
Rubin, M., & Hewstone, M. (1998). Social identity theory’s self-esteem hypothesis: A
review and some suggestions for clarification. Personality and Social Psychology
Bulletin, 2, 40–62.
Selman, R. L., & Byrne, D. F. (1974). A structural-developmental analysis of levels of
role-taking in middle childhood. Child Development, 45, 803–806.
Semaj, L. (1980). The development of racial evaluation and preference: A cognitive
approach. Journal of Black Psychology, 6, 59–79.
Stephan, W. G., & Rosenfield, D. (1979). Black self-rejection: Another look. Journal of
Educational Psychology, 71, 708–716.
Stephan, W. G., Ybarra, O., Martinez, C. M., Schwarzwald, J., & Tur-Kaspa, M.
(1998). Prejudice toward immigrants to Spain and Israel. Journal of Cross-Cultural
Psychology, 29, 559–576.
Stevenson, H. W., & Stevenson, N. G. (1960). Social interaction in an interracial
nursery school. Genetic Psychology Monographs, 61, 37–75.
Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categorisation and
intergroup behaviour. European Journal of Social Psychology, 1, 149–178.
Tajfel, H., Jahoda, G., Nemeth, C., Rim, Y., & Johnson, N. B. (1972). The devaluation
by children of their own national and ethnic group: Two case studies. British Jour-
nal of Social and Clinical Psychology, 11, 235–243.
Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In W. G.
Austin & S. Worchel (Eds.), The social psychology of intergroup relations. Monterey,
CA: Brooks/Cole.
Teplin, L. A. (1976). A comparison of racial/ethnic preferences among Anglo, Black
and Latino children. American Journal of Orthopsychiatry, 46, 702–709.
 Nesdale 245
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorisation theory. Oxford: Basil Blackwell.
Vaughan, G. M. (1963). Concept formation and the development of ethnic awareness.
Journal of Genetic Psychology, 103, 93–103.
Vaughan, G. M. (1964). The development of ethnic attitudes in New Zealand school
children. Genetic Psychology Monographs, 70, 135–175.
Vaughan, G. M. (1987). A social psychological model of ethnic identity development.
In J. S. Phinney & M. J. Rotheram (Eds.), Children’s ethnic socialisation. London:
SAGE Publishers.
Vaughan, G. M. (1988). The psychology of intergroup discrimination. New Zealand
Journal of Psychology, 17, 1–14.
Vaughan, G., Tajfel, H., & Williams, J. A. (1981). Bias in reward allocation in an
intergroup and an interpersonal context. Social Psychology Quarterly, 44, 37–42.
Vaughan, G. M., & Thompson, R. H. T. (1961). New Zealand children’s attitudes
toward Maoris. Journal of Abnormal Social Psychology, 62, 701–704.
Weiland, A., & Coughlin, R. (1979). Self-identification and preferences: A comparison
of white and Mexican-American first and third graders. Journal of Cross-Cultural
Psychology, 10, 356–365.
Williams, J. E., Best, D. L., & Boswell, D. A. (1975). The measurement of children’s
racial attitudes in the early school years. Child Development, 46, 494–500.
Williams, J. E., & Morland, J. K. (1976). Race, colour and the young child. Chapel Hill,
NC: University of North Carolina Press.
Williams, J. E., & Roberson, J. K. (1967). A method for assessing racial attitudes in
preschool children. Educational and Psychological Measurement, 27, 671–689.
Yee, M. D., & Brown, R. (1992). Self-evaluations and intergroup attitudes in children
aged three to nine. Child Development, 63, 619–629.
Yee, M. D., & Brown, R. (1994). The development of gender differentiation in young
children. British Journal of Social Psychology, 33, 183–196.
Zinser, O., Rich, M. C., & Bailey, R. C. (1981). Sharing behaviour and racial
preference in children. Motivation and Emotion, 5, 179–187.
9 The development and self-
regulation of intergroup
attitudes in children
Adam Rutland

From the age of five, children are exposed to the problems. Not mixing with
people from the other side of the community becomes the norm. It’s just not
the thing to do. It’s been like that for so many years that it’s normal now.
A Protestant from Belfast, Northern Ireland
(The Guardian, 4th January, 2002)

The above quotation suggests that at a surprisingly early age children may
form “self” and “other” representations in reference to social category
membership. It also suggests that when forming these intergroup attitudes,
children may be fully aware of the social norms of the prevailing context. The
aim of this chapter is to demonstrate that a comprehensive account of
children’s intergroup attitudes requires an understanding that children can
self-regulate their expression of intergroup bias in line with internalised
normative beliefs. Moreover, it will be argued that they do this strategically, to
present a positive image of themselves to valued members of their in-group.
The chapter starts with a short overview of the literature on children’s
developing intergroup attitudes and cognitive-developmental theory (CDT).
Research into national prejudice development will be presented to highlight
both empirical and conceptual problems with CDT. Next, social identity
theory (SIT) will be briefly discussed as an alternative account of prejudice
development. Finally some recent experiments will be outlined that demon-
strate how self-regulation plays an important role in the development of
intergroup attitudes.

CHILDREN’S INTERGROUP ATTITUDES AND COGNITIVE-

DEVELOPMENTAL THEORY

There is an abundance of psychological research suggesting that children

even as young as 3 years of age identify with social categories and express
clear affective preference for one rather than the other. For example, research
on ethnicity has shown that by the age of 3 or 4 years, children typically
248 Self-regulation of intergroup attitudes
express more positive attitudes towards their own ethnic group than towards
others (see reviews by Aboud, 1988; Aboud & Amato, 2001; Brown, 1995;
Cameron, Alvarez, Ruble, & Fuligni, 2001; Katz, 1976; Nesdale, 2001).
Negative intergroup attitudes with respect to gender are typically found
between 4 and 5 years of age (Bigler, 1995; Bigler & Liben, 1992; Powlishta,
1995a, 1995b; Powlishta, Serbin, Doyle, & White, 1994; Yee & Brown, 1994).
Studies have even shown intergroup bias in minimal or transitory groups at
around 5 and 6 years of age (e.g., Abrams, Rutland, Cameron, & Marques,
2003; Bigler, 1995; Bigler, Brown, & Markell, 2001; Bigler, Jones, & Lobliner,
1997; Durkin & Judge, 2001; Nesdale & Flesser, 2001; Spielman, 2000;
Vaughan, Tajfel, & Williams, 1981).
A common finding in the literature has been a “peaking” of childhood
prejudice around 5 to 7 years of age and then a decline in subsequent years.
Discovery of this developmental sequence was key to the formation of the
dominant theoretical account of children’s developing intergroup attitudes.
Cognitive-developmental theory (CDT) is probably the most well-known
account of how children’s intergroup attitudes develop (Aboud, 1988; Katz,
1976; Lambert & Klineberg, 1967). CDT argues that prejudice is caused by
information-processing errors due to young children’s poor cognitive ability
to perceive people of different racial groups in individualised terms
(e.g., friendly, hardworking). This theory argues that cognitively immature
young children are prone to prejudice because they cannot process multiple
classifications and attend simultaneously to two or more different perspec-
tives (i.e., decentre). Young children see the world only in bipolar terms and
cannot simultaneously process the internal and group-related qualities of an
individual. Thus they cannot see the similarities between individuals in differ-
ent groups and the differences between people within groups. Only with cog-
nitive development do children shift from social categorical to individuated
thinking and begin to make judgments in terms of unique interpersonal
rather than intergroup qualities.
CDT would predict that 5-year-old children should show high levels of
intergroup bias and that prejudice should “peak” at around 7 years of age
and then, with the acquisition of concrete operational thought, begin to
decline. There is a large body of evidence indicating that prejudice, especially
in the domain of ethnicity, is high among young children and then decreases
with age (see reviews, Aboud, 1988; Aboud & Amato, 2001). However, some
research suggests that this developmental sequence is not always evident.
There are various examples of research indicating that children in mid child-
hood and early adolescence still show intergroup bias (e.g., Abrams, 1985,
1989; Augoustinos & Rosewarne, 2001; Bennett, Lyons, Sani, & Barrett,
1998; Durkin & Houghton, 2000; Hoover & Fishbein, 1999; Rutland, 1999;
Tajfel, Nemeth, Jahoda, Campbell, & Johnson, 1970; Verkuyten, 2001). CDT
has difficulty accounting for prejudice development across all domains,
especially gender and nationality. For example, problematically for CDT,
there is evidence that girls are more likely than boys to maintain own-gender
 Rutland 249
bias attitudes irrespective of their age (Brown, 1995; Maccoby, 1988;
Powlishta et al., 1994; Yee & Brown, 1994). Furthermore, there are problems
using CDT to understand the intergroup attitudes of children from minority
or subordinate groups. Evidence suggests that children from dominant
groups almost always show strong in-group bias, whereas minority group
children show a much more varied pattern of responses, ranging from
out-group favouritism, through “fairness”, to in-group bias (Brown, 1995).
The findings of research into children’s national attitudes also raise
questions about the generality of CDT. For example, Tajfel et al. (1970)
conducted a large-scale cross-national study of children’s preference for their
own country. They found strong and significant evidence of intergroup bias
among 9- to 12-year-olds in England, Belgium, and Austria, though the
overall tendency to assign better-liked photographs to the national in-group
decreased with age. Interestingly, there were two exceptions to this general
pattern. In Louvain, regarding both the Belgian and Flemish assignments,
national intergroup bias showed a marked increase in age. However, in Glas-
gow the children of all ages showed little, if any, tendency to prefer their own
national group (i.e., “Scottish”). Tajfel, Jahoda, Nemeth, Rim, and Johnson
(1972) presented evidence suggesting that the lack of simple, unique, and
overtly salient national labels in the sociopolitical context of Belgium and
Scotland may explain the failure to find intergroup bias in young children.
They also contended that the “effects of social context on the attitudes of
children” (p. 239) was evidenced in the increase with age of national prefer-
ence among the Flemish but not the Scottish. Tajfel and colleagues argued
that self-denigrating stereotypes found among Scottish adults in the 1970s
(see Cheyne, 1970) might have affected children’s national attitudes. Indeed,
Tajfel and colleagues believed that they “provided evidence of the very high
sensitivity of young children to the more primitive aspects of the value
systems of their societies” (p. 243).

RECENT RESEARCH ON CHILDREN’S NATIONAL

INTERGROUP ATTITUDES

In 1995 I began a longitudinal research project to investigate the value of

CDT as an account of children’s intergroup attitudes. This project primarily
investigated changes in children’s national in-group bias and national preju-
dice over a 3-year period between 1995 and 1997 inclusive (see Rutland, 1999,
for the findings from data collection during 1995). Almost all the children
(aged 6 to 16 years) studied in 1995 were again interviewed in 1996 and also in
1997. Each year national in-group bias and prejudice was measured using the
same photograph evaluation task. This task involved the children evaluating
10 passport-style photographs showing people’s heads in two conditions:
with or without the national category label applied. The national labels were
British, American, German, Australian, and Russian.
250 Self-regulation of intergroup attitudes
The longitudinal data showed clearly that national in-group bias varied
over the 3 years of the project. The data from 1995 showed no evidence of
national in-group bias among the 6- and 8-year-olds. Bias was found only in
children over 10 years of age. However, there was a significant increase in
terms of national in-group bias across the two subsequent years of the proj-
ect. In 1996 and 1997 there was evidence of significant national in-group bias
in every age group. The children’s responses to informal questioning suggest
that these differences may be accounted for by children’s level of national
self-category salience. This is both the child’s awareness that they belong to
the national category and the degree to which they identify with the nation.
Regardless of age, levels of national self-category salience appeared higher in
1996 and 1997. This was especially true among the young children, and may
have reflected nationally significant events, such as the European Football
Championships (Euro ’96), held in the UK. These findings are in line with
previous research drawing upon SIT, which has shown that self-category
salience influences children’s intergroup attitudes (Bigler et al., 1997; Tajfel
et al., 1972) and, more specifically, that national identification affects national
in-group bias (Kinket & Verkuyten, 1999; Verkuyten, 2001). Evidence of
temporal variability in children’s national in-group bias poses an important
challenge to CDT. This theory would expect a stable age pattern in the devel-
opment children’s in-group bias across each year of the project, because
according to the theory, intergroup attitudes are primarily determined by the
child’s cognitive capacities. The data also clearly shows that national in-group
bias does not “peak” at 7 years of age and then decline, as predicted by
CDT. Indeed, it was evident that children maintained their levels of national
in-group bias well into adolescence.
Moreover, Rutland (1999) found in 1995 that children showed national
prejudice only towards the German out-group. No evidence of prejudice
towards the Americans, Australians, or Russians was found. Prejudice
towards Germans showed signs of appearing at 12 years of age and was
clearly evident among 14 and 16-year-olds. Again, in 1996 national prejudice
was only shown towards the Germans among the children above 13 years of
age. These findings too are incompatible with the predictions of CDT.
Importantly, the project also found evidence of temporal variability in
national prejudice towards the Germans. In 1997 the data showed no clear
sign of out-group derogation of the Germans. The most parsimonious
explanation of this finding is the lack of any significant event in 1997 to
highlight a contrast between the in-group and Germany. During 1997 there
was no significant event like the VE Day (Victory in Europe, i.e., the ending
of World War II in Europe) 50th anniversary in 1995 or Euro ’96, to focus
attention on Germany as a salient and distinctive out-group. This temporal
shift in prejudice towards Germans raises doubts about the explanatory
power of CDT and suggests that contextual factors may play an important
part in the onset of prejudicial beliefs among children.
A recent study by Fyfe and Rutland (2000), like Tajfel et al., (1970), found
 Rutland 251
cross-national differences in children’s national intergroup attitudes. The
nations used within this study were chosen because they varied in terms of
group status. Contrary to CDT, social identity theory (SIT) would predict
that children from high-status groups should show greater intergroup bias
than those from low-status groups because their group membership should
allow for greater enhancement of self-esteem (Bigler et al., 2001; Brown,
1984; Doosje, Ellemers, & Spears, 1999; Nesdale & Flesser, 2001; Yee &
Brown, 1994). Seventy-two Maltese children (i.e., from a “low”-status
country) and the same number of Scottish children (i.e., from a relatively
“high”-status country) were included in the study. These children were aged
between 6 and 16 years. A picture preference task was used to assess attitudes
towards the in-group and the English, Germans, French, and Italians.
The Maltese showed no evidence of national in-group bias or out-group
derogation. In contrast, the Scottish children showed significant national
in-group bias, some evidence of negativity towards French and Maltese at 6
to 8 years of age, and signs of negativity towards the English and Germans
at 14 to 16 years of age. Like the early work by Tajfel and colleagues, this
study suggests children’s national intergroup attitudes are very sensitive to
contextual factors.
The findings of the studies described above raise doubts about the general-
ity of CDT. The “dip” predicted by CDT in children’s negative intergroup
attitudes was not evident in these studies of national intergroup attitudes.
Rather, the children’s attitudes towards countries showed both contextual and
temporal variability. There was evidence of national in-group bias among
young children in both England and Scotland, while it was absent among
Maltese children. National prejudice did manifest itself among English and
Scottish children, though this was often particular to a salient out-group and
showed temporal flexibility. In contrast, no evidence of national prejudice
was found among Maltese children.

SOCIAL IDENTITY THEORY

The limitations associated with CDT have encouraged some researchers to

turn to social identity theory (SIT: Tajfel & Turner, 1979, 1986) as an
alternative account of prejudice development (Nesdale, 2001). SIT has been
described in sufficient detail elsewhere not to require a further full exposition
here (see Chapter 1 by Bennett & Sani, and Chapter 8 by Nesdale; also
Brown, 2000; Capozza & Brown, 2000; Ellemers, Spears, & Doosje, 1999).
Despite its prominence within the social-psychological literature, SIT says
nothing about age trends in children’s intergroup attitudes and consequently
is unable to account for age effects in this field when they are found. Children
clearly undergo significant age-linked changes in their cognitive, linguistic,
and social abilities and these changes are likely to have a significant bearing
on their intergroup attitudes. SIT, with its emphasis on motivational concerns
252 Self-regulation of intergroup attitudes
and the social context, is silent on how these changes may be associated
with particular age trends in children’s intergroup attitudes. There is no
mechanism or psychological process within SIT to explain why children’s
attitudes can change as they become older. Specifically, these theories have a
problem explaining the common finding that ethnic intergroup bias is strong
among young children and then decreases with age (Aboud, 1988; Doyle &
Aboud, 1995).

SELF-REGULATION AND CHILDREN’S

INTERGROUP ATTITUDES

No one theory to date has offered a comprehensive explanation of children’s

developing intergroup attitudes. CDT has provided a clear account of the
developmental trends commonly found in the domain of ethnic prejudice –
namely the well-documented “peaking” of prejudice at around 5 to 7 years
of age and then the gradual decline into late childhood (see Aboud, 1988;
Brown, 1995; Katz, 1976). However, CDT faces problems when attempting
to explain children’s intergroup attitudes in other domains (e.g., nationality
and gender). This nevertheless does not negate the fact that clear develop-
mental patterns, especially in ethnic prejudice, are often found in the litera-
ture on children’s intergroup attitudes, and these need explanation. SIT is not
a developmental theory and therefore make no specific predictions regarding
developmental trends in children’s intergroup attitudes.
The findings of previous research on national and ethnic prejudice may
become compatible if it is accepted that the developmental path of prejudice
may be dependent on children’s ability to self-regulate their expression of
prejudice in accordance with normative beliefs. Children may gradually learn
with age to associate particular labels (e.g., nasty, friendly) with certain types
of social categories (e.g., blacks, whites, Germans, French). However the
degree to which they demonstrate explicit prejudice towards these categories
may well be moderated by their ability to critically evaluate the legitimacy
and acceptability of these implicit associations. This process of critical evalu-
ation could be influenced to a large degree by internalised normative beliefs
about the expression of particular views. The expression of negative attitudes,
then, may vary as a function of both age and domain. Thus, given the recog-
nition that children can regulate their expression of prejudice, it is empirically
conceivable that sometimes children above 7 to 9 years of age will show low
prejudice (i.e., with regard to ethnic groups) and also high prejudice (i.e., with
regard to national groups).
This alternative account of development trends in children’s intergroup
attitudes was hinted at when Brown (1995) stated, “one might add, too,
though this is not much emphasized in Aboud’s account, that as children
mature they become more sensitive to norms of adult society and hence more
aware of the social undesirability of expressing certain kinds of prejudice
 Rutland 253
too overtly” (p. 155). It has been argued before that the internalisation of
normative beliefs regarding legitimate and admissible forms of prejudice and
stereotyping might help explain differences in the expression of stereotypical
thoughts (e.g., Macrae, Bodenhausen, & Milne, 1997). Rutland (1999)
explained the contradiction between his results on national prejudice develop-
ment and previous research on ethnic prejudice in terms of the divergent
norms in our society regarding the legitimacy of expressing national and
ethnic prejudice. It is fair to say that in most societies, prejudice is viewed
negatively, as irrational and uninformed thought. Among most people ethnic
prejudice is certainly seen as unreasonable, but with national prejudice this is
not always the case. Sometimes it is seen as a tolerable form of expression
among the majority (Billig, 1991). It seems reasonable, then, to suggest that
older children and adolescents may show reduced ethnic prejudice due to the
influence of social norms regarding the illegitimacy of ethnic prejudice. In
contrast, social norms surrounding national prejudice might actually encour-
age more prejudice. Indeed, research involving adults has clearly shown that
manipulation of in-group social norms can increase levels of intergroup dis-
crimination (e.g., Jetten, Spears, & Manstead, 1996, 1997; Long & Manstead,
1997; Rutland & Brown, 2001).
There is evidence that children can actively engage in self-presentation and
present normative intergroup behaviour to win positive evaluations from
their in-group. Jahoda, Thomson, and Bhatt (1972) investigated ethnic iden-
tity and preferences among 6- to 10-year-old Asian children in Glasgow. In
their first study they found that the children’s host community powerfully
influenced Asian children as they showed preferences towards the white out-
group. However, in their second study the ethnicity of the experimenter was
changed as they used an Asian psychologist to test the children. This had a
significant effect on the Asian children’s responses. They no longer showed
out-group favouritism; rather, their preferences moved in the direction of
Asian cultural values and they showed significant in-group bias. Jahoda and
colleagues concluded that it is difficult to ascertain the “real” values of these
children: they “may be in a state of somewhat precarious balance, fraught
with potential conflict; they can only avoid this by expressing the values that
appear to them situationally appropriate” (p. 31). A more recent study also
showed experimenter effects on children’s intergroup attitudes. Pedersen,
Walker, and Glass (1999) found in-group preferences were higher in 5- to
12-year-old Aboriginal-Australian children using an in-group Aboriginal
compared to an out-group white experimenter.
Recently, some social identity theorists researching with adults have
argued that conscious, strategic self-management may override automatic
responses in certain contexts and drive intergroup attitudes, especially when
the audience or social context requires the avoidance of negative, and the
promotion of positive, identities (Abrams, 1990, 1994; Abrams & Brown,
1989; Barreto & Ellemers, 2000; Noel, Wann, & Branscombe, 1995; Postmes,
Branscombe, Spears, & Young, 1999; Reicher, Spears, & Postmes, 1995;
254 Self-regulation of intergroup attitudes
Spears, 2001). This research suggests that the exclusive use of explicit
measures when investigating intergroup attitudes allows participants to
strategically claim positive and valued identities (e.g., a fair-minded person),
while downplaying negative identities (e.g., an irrational bigot).
Research in developmental psychology indicates that even children as
young as 4 years of age can actively engage in self-presentational efforts to
win positive identities and evaluations from the in-group when they are highly
motivated to make good impressions. Banerjee and Lintern (2000) found that
4- to 6-year-old boys produced more gender-stereotypical self-descriptions
when in front of a group of same-sex peers than when alone. There was no
evidence of a peer audience effect on the self-descriptions of 4- to 8-year-old
girls. This finding was not unexpected, given that research has shown that
girls demonstrate less gender-typing than boys (e.g., Huston, 1985; Serbin,
Conner, Burchardt, & Citron, 1979). However, as Banerjee and Lintern sug-
gest, it implies that peer groups of males produce more conformity to in-
group norms than do females. The fact that peer group effects were strongest
for young boys, who showed high levels of gender-typing in the baseline
condition, suggests males are concerned about social evaluation because they
wish to appear consistent with the norms of the peer group. These results and
those from other studies (e.g., Banerjee & Yuill, 1999a, 1999b; Bennett &
Yeeles, 1990) indicate that school-age children have the capacity to under-
stand self-presentational motives and, by approximately 8 years of age, they
can provide spontaneous explanations for complex self-presentational
behaviour.

IMPLICIT MEASURES OF CHILDREN’S

INTERGROUP ATTITUDES

One way to examine whether children regulate the expression of prejudice is

to use methods that limit the possibility of their giving socially desirable
responses. Researchers have begun to countenance the possibility that reduc-
tions of prejudice with age may reflect “greater sensitivity to the social
appropriateness of responses” (Powlishta et al., 1994, p. 533). Indeed,
Powlishta and her colleagues suggest that to reduce the potential impact of
social desirability responding, future studies should adopt more subtle meas-
ures of prejudice that are less dependent on conscious awareness. Nesdale
(2001) made a similar point when he reviewed the existing measures typically
used in research on childhood prejudice. He stated that the “shared weakness
of each of the preceding measurement techniques concerns their transpar-
ency of intent, even to young children, with the resulting problem that each
invites children to construct responses which they consider to be more
socially acceptable” (p. 62). Researchers are beginning to realise that new
methodologies are required to minimise social desirability effects in children.
Indeed, Nesdale suggests the transparency problems can be handled by using
 Rutland 255
psychologically relevant measures of which children are unaware, namely
implicit rather than explicit measures of bias. Though such measures are
becoming more common in social psychology (see Devine, 2001; Greenwald
& Banaji, 1995; Maass, Castelli, & Arcuri, 2000), they have rarely been used
with children.
Most research on childhood prejudice has used direct measures of atti-
tudes, which by implication make the assumption that attitudes operate in a
conscious mode and are exemplified by traditional self-report measures.
However, recent work within social psychology has established that attitudes
can be activated automatically and outside of conscious attention (e.g.,
Bargh, Chaiken, Govender, & Pratto, 1992). This can result in judgments
occurring without awareness of the causation. Greenwald and Banaji (1995)
defined implicit attitudes as “introspectively unidentified (or inaccurately
identified) traces of past experience that mediate favorable or unfavorable
feeling, thought, or action toward social objects” (p. 8). Intuitively one might
expect implicit prejudice, as assessed in response latency paradigms, and
explicit self-reported prejudice, to be based on the same experiences and
socialisation. Thus they should be directly related, but research typically does
not support this expectation; on the contrary: Dissociations are common
(Banaji & Greenwald, 1995; Dovidio, Kawakami, Johnson, Johnson, &
Howard, 1997). Differences between implicit and explicit measures are espe-
cially likely to be observed for socially sensitive issues and particularly for
racial attitudes (Greenwald, McGee, & Schwartz, 1998). Conceivably,
differences between response latency and self-report measures may reflect the
distinction between activation and application. The presence of an attitude
object may automatically activate an association evaluation from memory,
which may be applied in subsequent judgments (Dovidio et al., 1997).

DISSOCIATION BETWEEN IMPLICIT AND

EXPLICIT MEASURES

Evidence of a dissociation between response latency measures of implicit bias

and self-reported explicit measures of bias would suggest that children are
engaging in self-regulation when expressing their explicit attitudes. Children
should gradually become aware of particular implicit associations surround-
ing social categories, but whether these associations are endorsed and applied
will depend on children’s self-regulation of explicit prejudice in line with
normative beliefs. Therefore, when social norms prohibit the articulation of
explicit bias we should find a weak relationship between measures of implicit
and explicit prejudice. Rutland, Morrison, and Arnold-Dorman (2001) exam-
ined the relationship between implicit and explicit measures in two studies of
young females’ body image bias. Among young females there is evidence of
a decrease in explicit body image bias, with age. Powlishta et al. (1994) found
females showed less bias against overweight body images with increasing age.
256 Self-regulation of intergroup attitudes
This finding suggests that, with age, females become increasingly aware of the
unacceptability of expressing blatant body image bias, even though they may
still, at some level, subscribe to an image of beauty that does not involve an
overweight body image (e.g., Balaam & Haslam, 1998; Muth & Cash, 1997).
Given these findings, it seems reasonable to expect dissociation between
measures of implicit and explicit body image bias.
The first study examined 60 females aged between 8 and 21 years. They
were presented with an explicit and an implicit task. The explicit measure was
a trait attribution task similar to the one utilised by Powlishta et al. (1994).
The participants had to indicate which of two female silhouettes, one slim
and one overweight, was more likely to be characterised by each of a list of 12
positive and 8 negative traits. The implicit measure was the Stroop colour-
naming task, a task that has previously been used to measure adults’ process-
ing of weight- and shape-related words (e.g., Cooper & Fairburn, 1992).
Each participant was presented with three computer screens of words that
constituted three conditions (neutral, colour, and weight). These screens con-
tained five word columns with the same five words in each column. The
columns contained either neutral words (sit, dare, object, tower, wool), colour
words (orange, red, blue, green, brown), or weight/body shape words (fat,
diet, thighs, cakes, hips). The words were written in different colour ink and
the participants were asked to name the colour of all the words on each
screen. The ordering of the ink colour and the words on each screen were
randomised. Response times for correct completion of each screen were
recorded. It was predicted that participants would be slower responding to
the weight word screen than the neutral word screen. This finding would
indicate selective processing of information relating to weight words among
our female participants.
The findings showed a noticeable drop in the females’ explicit bias with age.
There was a significant decrease with age in the number of positive attributes
assigned to the slim silhouette, whereas the number of positive attributions to
the overweight silhouette increased with age. Negative trait attributions to the
slim silhouette increased with age, while the overweight silhouette received
significant fewer negative attributions with age. Post hoc analysis revealed
across all attributions that body image bias decreased significantly between
8–9 years and 14–15 years of age.
In contrast, the implicit measure showed the Stroop effect in all age groups.
The females’ response times at all ages were significantly faster for the neutral
words compared to the weight-salient and colour words. There was no evi-
dence of an equivalent decrease in this implicit measure with age, and these
findings were replicated in the second study. Together these results suggest
the females had a significant subconscious disturbance or interference regard-
ing weight. Though this indicates that the females had an implicit difficulty
with weight issues it does not necessarily suggest that they were implicitly
biased against overweight people. Indeed, the Stroop naming task has been
traditionally used within clinical psychology more as a measure of implicit
 Rutland 257
self-body image disturbance rather than implicit bias towards body shape
(e.g., Cooper & Fairburn, 1992; Cooper, Anastasiades, & Fairburn, 1992).

THE IMPLICIT ASSOCIATION TEST

The second study, therefore, also included an implicit method known to

measure implicit attitudinal bias directly. This was the Implicit Association
Test (IAT) developed by Greenwald et al. (1998). The IAT was devised to
measure automatic concept-attribute associations and an underlying assump-
tion of the test is that strongly associated (compatible) attribute-concept
pairs should be easier to classify together than weakly associated or opposed
(incompatible) attribute-concept pairs. The IAT measures implicit bias in the
participants by presenting them with a series of words on a computer screen.
They have to categorise these words (as compatible or incompatible) as
quickly as possible by pressing a left or right key on a keyboard.
The IAT consisted of a sequence of five blocks, which together assessed the
association between a target concept and an attribute dimension. The pro-
cedure started with an introduction to the target concept block. In this block
the participants were required to categorise words that were recognisable as
either coming from one of two concepts: slim (e.g., slender or toned) or
overweight (obese or fat). The second block involved the introduction of the
attribute dimension. The children were presented with either positive or nega-
tive adjectives and asked to categorise these words as pleasant versus
unpleasant in meaning. After this introduction to the target concepts and to
the attribute dimension, the two were superimposed in the third block. The
stimuli for target and attribute discriminations appeared on alternate trials.
This was the intuitive block, as the target slim words were paired with positive
words and the target overweight words were paired with the negative words.
In the fourth block, the participants learnt a reversal of response assignments
for the target concepts: They were presented with the same stimulus as in the
first block except they responded to slim and overweight words using the
opposite keys (i.e., left or right) to those used in the first block. Then in
the fifth and final counterintuitive block, the target concepts were reversed
and combined with the same attribute discriminations as in third block. This
meant that the target slim words were now paired with negative words and the
target overweight words were paired with the positive words.
It was found that the target concepts (“slim” and overweight) were differ-
entially associated with the positive and negative attributes. The participants
in all age groups recorded quicker response times in the intuitive block
compared to the counterintuitive block (i.e., an IAT effect). The concept
“slim” was seen as more compatible with positive adjectives than the concept
“overweight”; and vice versa for negative adjectives. Unlike explicit body
image bias, the IAT effect did not decrease with age. In addition, there was
not a significant correlation between the explicit and implicit measures. These
258 Self-regulation of intergroup attitudes
two studies show that implicit and explicit measures of body image bias
follow different developmental paths. Dissociation between these two types
of measure suggests that the participants were sensitive to social desirability
and normative concerns. This is evidenced by the fact that with increasing age
they were unwilling to express explicit or blatant body image bias, even
though they still showed signs of implicit bias regarding an overweight body
image.

PUBLIC SELF-FOCUS AND CHILDREN’S ETHNIC

INTERGROUP ATTITUDES

The argument that children’s intergroup attitudes are dependent on their

ability to engage in self-presentation and regulate their expression of preju-
dice in accordance with normative beliefs can also be investigated through
studies that manipulate public self-focus. Public self-focus involves attention
to one’s self-portrayal (Scheier & Carver, 1981), namely, to how significant
others may perceive the self in a specific context. Abrams (1990) suggests that
when people focus on the public self they take up general standards of
socially legitimate behaviour (e.g., equity norms) and show intragroup
compliance. This means that when attention is directed towards managing an
appropriate public impression, behaviour may follow the specific norms of
the audience or, if these are unclear, general social norms (e.g., to please the
experimenter).
A recent study by Rutland, Cameron, Milne, and McGeorge (2002)
investigated directly whether heightened public self-focus in children will typ-
ically reduce explicit ethnic bias in compliance with internalised normative
beliefs. Evidence of self-presentation would be apparent if the children
showed significantly lower explicit ethnic bias when public self-focus is high
rather than low. However, this effect was predicted only among young
children, since research (e.g., Bennett & Yeeles, 1990) indicates that children
above approximately 8 years of age spontaneously understand and engage in
self-presentational behaviour when they are highly motivated to make good
impressions or claim a positive identity. This study also examined the devel-
opmental paths of explicit and implicit ethnic bias through the inclusion of
the modified pictorial version of the IAT (Greenwald et al., 1998) as an
implicit measure of intergroup attitude. No equivalent drop in implicit ethnic
bias was expected with the public self-focus manipulation since implicit atti-
tudes should be automatically activated outside conscious self-control
(Greenwald & Banaji, 1995).
The study involved 162 white children with equal numbers in three age
groups: 6–8, 10–12, and 14–16 years. Initially, a preliminary study determined
the prevailing normative context in a similar cohort of 118 children. These
14- to 16-year-olds were asked to rate on a 9-point scale (the higher the score
the more inappropriate) how appropriate they thought it was in their society
 Rutland 259
to judge someone on the basis of their social group. Thirty-five social groups
were used (e.g., blacks, Asians, criminals, French, Germans, and whites). The
mean score for all ethnic groups combined was 7.23, significantly higher than
the composite score for all national groups (M = 6.75; p < .01) and criminals
(M = 4.38; p < .001). These findings suggest that the children thought it was
extremely inappropriate to judge someone on the basis of their ethnic group.
Following this, our sample was presented with two social exclusion
vignettes. One was used with boys and involved two black boys being
excluded by white boys from a five-a-side soccer game because of their eth-
nicity. The other vignette was presented to girls and showed two black girls
being excluded by white girls from eating their school lunch during a break,
again because of their ethnicity. These vignettes were used to evaluate
whether children had internalised the norm that ethnic bias was inappropri-
ate. After the vignettes the children were asked how bad they thought the
white children were. They had to answer on a 4-point scale (1= OK, 2 = bad, 3
= very bad, and 4 = very very bad). Overall only 3% of the children answered
“OK” and 65% said “very very bad”. There were significant age differences in
their answers to the question. The 14–16 (M = 3.66) and 10–12 (M = 3.57) age
groups gave more negative answers compared to the 6–8 age group (M =
3.27). Nevertheless these means indicate that children in all age groups had to
a certain degree internalised the norm surrounding ethnic prejudice, although
the internalisation process had clearly progressed further in the older
children.
The explicit measure in the main study was a trait attribution task similar
to the one utilised by Powlishta et al. (1994). The participants had to indicate
which of two social groups (blacks or whites) was more likely to display a list
of 10 positive and 10 negative traits. A positive attribution bias score was
computed by subtracting the number of positive traits assigned to blacks
from the number given to whites: The higher the score, the more in-group
bias. A negative attribution bias score was obtained by subtracting the num-
ber of negative traits assigned to whites from the number given to blacks: The
higher the score, the more negative out-group bias. Public self-focus was
manipulated using a video camera (see Macrae et al., 1997; Wicklund &
Duval, 1971). In the high public self-focus condition the children completed
the explicit and implicit measures when the video camera was turned “on”.
The children were told that a video recording was being taken to keep a
record of the testing session and that it may be shown to other adults. The
video camera was turned “off” in the low public self-focus condition.
In the low public self-focus condition there was the usual decline in ethnic
positive bias with age. The 6- to 8 year-olds showed significantly more posi-
tive bias than the 10- to 12 and 14- to 16-year-olds. However, there was no
significant decline in ethnic positive bias with age in the high public self-focus
condition. The age groups showed similar low levels of positive bias.
The positive bias scores of the 6- to 8-year-olds were significantly above the
midpoint in the low public self-focus condition but not in the high public
260 Self-regulation of intergroup attitudes
self-focus condition. In relation to the ethnic negative bias score there was a
significant age × public self-focus interaction. Again there was the expected
decline in negative bias with age in the low public self-focus condition. The
6- to 8-year-olds showed significantly more negative bias than the 10- to 12
and 14- to 16-year-olds. In contrast, when public self-focus was high, each age
group showed similar levels of ethnic negative bias. The 6- to 8-year-old
children’s negative bias scores in the low public self-focus condition could
be seen as ethnic prejudice since they were significantly above the midpoint of
the scale.
The IAT scores showed significant implicit ethnic bias amongst all chil-
dren. All age groups produced faster reaction times in the intuitive block
compared to the counterintuitive block. That is, children were quicker to pair
positive attributes with white faces and negative attributes with black faces
than visa versa. The public self-focus manipulation had no effect on the IAT
scores and correlations between the explicit measures and the IAT scores
were nonsignificant.
These findings were in line with our expectations. The 6- to 8-year-old
children showed more positive ethnic intergroup attitudes when public self-
focus was high rather than low. The usual decline in explicit ethnic bias was
only found when public self-focus was low. These results suggest that the
young children were inhibiting their explicit ethnic bias according to norma-
tive beliefs surrounding ethnic prejudice expression when public self-focus
was high. That is, they were engaging in self-presentation. The older chil-
dren’s ethnic intergroup attitudes were unaffected by the public self-focus
manipulation. The simplest explanation for this finding is that the older chil-
dren were spontaneously enacting self-presentation and bias regulation even
in the low public self-focus condition, whereas the young children required
high public self-focus in order to prompt a self-presentational strategy. The
differences between the age groups on the vignette tasks measuring the
internalisation of normative beliefs suggests that the older children had a
firmer grasp of the normative context. Nevertheless, the vast majority of
the younger children still considered the behaviour of the white children
in the social exclusion stories to be “very very bad”. This implies that age
differences in responses to the public self-focus manipulation may have little
to do with whether norms were internalised. Rather these differences may
be better explained by young children’s inability to engage in spontaneous
self-presentation in this domain of ethnic intergroup attitudes.

CONCLUSION

At present no one theory can fully explain the development of inter-

group attitudes in children. Each theory has its limitations. Cognitive-
developmental theory has particular problems explaining variations in
prejudice development over different domains (e.g., gender and nationality)
 Rutland 261
and why it is that prejudice is sometimes still strongly evident in mid child-
hood and adolescence. In contrast, SIT makes no attempt to provide an
account of clear developmental trends in ethnic attitudes throughout child-
hood. This theory specifies no mechanisms or psychological processes that
can explain why, for example, young children show strong ethnic prejudice
that then declines in mid childhood and into adolescence. The research
reviewed at the end of this chapter suggests a resolution to many of the
anomalies in the literature on childhood prejudice. This requires a recogni-
tion that children can self-regulate their expression of explicit prejudice in
accordance with internalised normative beliefs. Arguably, then, children
can be seen as social tacticians who strategically express their intergroup
attitudes through a process of effortful on-line construction in context.
Their aim may well be to present a positive image to valued others
within their collectives and so to win a positive self-identity. Indeed, the
strategic dimension of the child’s expression of prejudice may be relational
to the ideological and normative significance of the particular social
category. More research is undoubtedly needed to further validate this
claim.

REFERENCES

Aboud, F. E. (1988). Children and prejudice. Oxford: Blackwell.

Aboud, F. E., & Amato, M. (2001). Developmental and socialization influences on
intergroup bias. In R. Brown & S. L. Gaertner (Eds.), Blackwell handbook of social
psychology: Intergroup processes (pp. 65–85). Oxford: Blackwell.
Abrams, D. (1985). Focus of attention in minimal intergroup discrimination. British
Journal of Social Psychology, 24, 65–74.
Abrams, D. (1989). Differential association: Social developments in gender identifica-
tion during adolescence. In S. Skevington & D. Baker (Eds.), The social identity of
women (pp. 59–83). London: SAGE Publications.
Abrams, D. (1990). The self regulation of group behaviour: An integration of self
awareness and social identity theory. In D. Abrams & M. A. Hogg (Eds.),
Social identity theory: Constructive and critical advances (pp. 89–112). London:
Harvester Wheatsheaf.
Abrams, D. (1994). Social self-regulation. Personality and Social Psychology Bulletin,
20, 473–484.
Abrams, D., & Brown, R. J. (1989). Self-consciousness and social identity:
Self-regulation as a group member. Social Psychology Quarterly, 52, 311–318.
Abrams, D., Rutland, A., Cameron, L., & Marques, J. (2003). The development
of subjective group dynamics: When in-group bios gets specific. British Journal of
Developmental Psychology, 21, 155–176.
Augoustinos, M., & Rosewarne, D. L. (2001). Stereotype knowledge and prejudice in
children. British Journal of Developmental Psychology, 19, 143–156.
Balaam, B. J., & Haslam, A. S. (1998). A closer look at the role of social influence in
the development of attitudes to eating. Journal of Community and Applied Social
Psychology, 8, 195–212.
262 Self-regulation of intergroup attitudes
Banaji, M. R., & Greenwald, A. G. (1995). Implicit gender stereotyping in judgments
of fame. Journal of Personality and Social Psychology, 68, 181–198.
Banerjee, R., & Lintern, V. (2000). Boys will be boys: The effect of social evaluation
concerns on gender typing. Social Development, 9, 397–408.
Banerjee, R., & Yuill, N. (1999a). Children’s explanations of self-presentational
behaviour. European Journal of Social Psychology, 29, 105–111.
Banerjee, R., & Yuill, N. (1999b). Children’s understanding of self-presentational
display rules: Associations with mental-state understanding. British Journal of
Developmental Psychology, 17, 111–124.
Bargh, J. A., Chaiken, S., Govender, R., & Pratto, F. (1992). The generality of the
automatic activation effect. Journal of Personality and Social Pyschology, 62,
893–912.
Barreto, M., & Ellemers, N. (2000). You can’t always do what you want: Social
identity and self-presentational determinants of the choice to work for a low status
group. Personality and Social Psychology Bulletin, 26, 891–906.
Bennett, M., Lyons, E., Sani, F., & Barrett, M. (1998). Children’s subjective
identification with the group and in-group favoritism. Developmental Psychology,
34, 902–909.
Bennett, M., & Yeeles, C. (1990). Children’s understanding of the self-presentational
strategies of ingratiation and self-promotion. European Journal of Social Psychology,
20, 455–461.
Bigler, R. S. (1995). The role of classification skill in moderating environmental
influences on children’s gender stereotyping: A study of the functional use of
gender in the classroom. Child Development, 66, 1072–1087.
Bigler, R. S., Brown, C. S., & Markell, M. (2001). When groups are not created equal:
Effects of group status on the formation of intergroup attitudes in children. Child
Development, 72, 1151–1162.
Bigler, R. S., Jones, L. C., & Lobliner, D. B. (1997). Social categorization and the
formation of intergroup attitudes in children. Child Development, 68, 530–543.
Bigler, R. S., & Liben, L. S. (1992). Cognitive mechanisms in children’s gender stereo-
typing: Theoretical and educational implications of a cognitive-based intervention.
Child Development, 63, 1351–1363.
Billig, M. (1991). Ideology and opinions: Studies in rhetorical psychology. London:
SAGE Publications.
Brown, R. J. (1984). The effects of intergroup similarity and cooperative versus
competitive orientation on intergroup discrimination. British Journal of Social
Psychology, 23, 21–33.
Brown, R. J. (1995). Prejudice: Its social psychology. Oxford: Blackwell.
Brown, R. J. (2000). Social identity theory: Past achievements, current problems and
future challenges. European Journal of Social Psychology, 30, 745–778.
Cameron, J. A., Alvarez, J. M., Ruble, D., & Fuligni, A. J. (2001). Children’s lay
theories about in-groups and out-groups: Reconceptualizing research on prejudice.
Personality and Social Psychology Review, 5, 118–128.
Capozza, D., & Brown, R. (Eds.) (2000). Social identity processes: Trends in theory and
research. London: SAGE Publications.
Cheyne, W. N. (1970). Stereotyped reactions to speakers with Scottish and
English regional accents. British Journal of Social and Clinical Psychology, 9,
77–79.
Cooper, M. J., Anastasiades, P., & Fairburn, C. G. (1992). Selective processing of
 Rutland 263
eating, shape and weight-related words in persons with bulimia nervosa. Journal of
Abnormal Psychology, 101, 352–355.
Cooper, M. J., & Fairburn, C. G. (1992). Selective processing of eating, weight and
shape related words in patients with eating disorders and dieters. British Journal of
Clinical Psychology, 31, 363–365.
Devine, P. G. (2001). Implicit prejudice and stereotyping: How automatic are they?
Introduction to the special section. Journal of Personality and Social Psychology, 81,
757–759.
Doosje, B., Ellemers, N., & Spears, R. (1999). Commitment and intergroup behaviour.
In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social identity, context, commitment
and content (pp. 84–106). Oxford: Blackwell.
Dovidio, J. F., Kawakami, K., Johnson, C., Johnson, B., & Howard, A. (1997). On the
nature of prejudice: Automatic and controlled processes. Journal of Experimental
Social Psychology, 33, 510–540.
Doyle, A., & Aboud, F. (1995). A longitudinal study of white children’s racial
prejudice as a social-cognitive development. Merrill-Palmer Quarterly, 41,
209–228.
Durkin, K., & Houghton, S. (2000). Children’s and adolescents’ stereotypes of
tattooed people as delinquent. Legal and Criminological Psychology, 5, 153–164.
Durkin, K., & Judge, J. (2001). Effects of language and social behaviour on children’s
reactions to foreign people in television. British Journal of Developmental
Psychology, 19, 597–612.
Ellemers, N., Spears, R., & Doosje, B. (Eds.) (1999). Social identity, context,
commitment and content. Oxford: Blackwell.
Fyfe, C., & Rutland, A. (2000, September). A comparative study of national prejudice
and in-group bias in Scottish and Maltese children. Paper presented at the British
Psychological Society (BPS) Social Psychology Section Annual Conference,
Nottingham Trent University.
Greenwald, A. D., & Banaji, M. R. (1995). Implicit social cognition: Attitudes, self-
esteem, and stereotypes. Psychological Review, 102, 4–27.
Greenwald, A. D., McGee, D. E., & J. L. K. Schwartz (1998). Measuring individual
differences in implicit cognition: The implicit association test. Journal of Personality
and Social Psychology, 74, 1464–1480.
Hoover, R., & Fishbein, H. D. (1999). The development of prejudice and sex role
stereotyping in white young adolescents and white young adults. Journal of Applied
Developmental Psychology, 20, 431–448.
Huston, A. C. (1985). The development of sex typing: Themes from recent research.
Developmental Review, 5, 1–17.
Jahoda, G., Thomson, S. S., & Bhatt, S. (1972). Ethnic identity and preferences among
Asian immigrant children in Glasgow: A replicated study. European Journal of
Social Psychology, 2, 19–32.
Jetten, J., Spears, R., & Manstead, A. S. R. (1996). Intergroup norms and intergroup
discrimination: Distinctive self-categorization and social identity effects. Journal of
Personality and Social Psychology, 71, 1222–1223.
Jetten, J., Spears, R. & Manstead, A. S. R. (1997). Strength of identification and
intergroup differentiation: The influence of group norms. European Journal of
Social Psychology, 27, 603–609.
Katz, P. A. (1976). The acquisition of racial attitudes in children. In P. A. Katz (Ed.),
Towards the elimination of racism (pp. 125–154). New York: Pergamon.
264 Self-regulation of intergroup attitudes
Kinket, B., & Verkuyten, M. (1999). Intergroup evaluations and social context: A
multilevel approach. European Journal of Social Psychology, 29, 219–237.
Lambert, W. E., & Klineberg, O. (1967). Children’s views of foreign peoples: A
cross-national study. New York: Appleton-Century-Crofts.
Long, K. M., & Manstead, A. S. R. (1997). Group immersion and intergroup differen-
tiation: Contextual shifts in categorization. British Journal of Social Psychology, 36,
291–303.
Maass, A., Castelli, L., & Arcuri, L. (2000). Measuring prejudice: Implicit versus
explicit techniques. In D. Capozza & R. Brown (Eds.), Social identity processes:
Trends in theory and research. London: SAGE Publications.
Maccoby, E. (1988). Gender as a social category. Developmental Psychology, 24,
755–765.
Macrae, C. N., Bodenhausen, G. V., & Milne, A. B. (1997). Saying no to unwanted
thoughts: Self-focus and the regulation of mental life. Journal of Personality and
Social Psychology, 74, 578–589.
Muth, J. L., & Cash, T. F. (1997). Body-image attitudes: What differences does gender
make? Journal of Applied Social Psychology, 27, 1436–1452.
Nesdale, D. (2001). Development of prejudice in children. In M. Augoustinos & K. J.
Reynolds (Eds.), Understanding prejudice, racism and social conflict (pp. 57–72).
London: SAGE Publications.
Nesdale, D., & Flesser, D. (2001). Social identity and the development of children’s
group attitudes. Child Development, 72, 506–517.
Noel, J. G., Wann, D. L., & Branscombe, N. R. (1995). Peripheral in-group
membership status and public negativity toward out-groups. Journal of Personality
and Social Psychology, 68, 127–137.
Pedersen, A., Walker, I., & Glass, C. (1999). Experimenter effects on in-group
preference and self-concept of urban Aboriginal children. Australian Journal of
Psychology, 51, 82–89.
Postmes, T., Branscombe, N., Spears, R., & Young, H. (1999). Comparative processes
in personal and group judgments: Resolving the discrepancy. Journal of Personality
and Social Psychology, 76, 320–338.
Powlishta, K. (1995a). Intergroup processes in childhood: Social categorization and
sex role development. Developmental Psychology, 31, 781–788.
Powlishta, K. (1995b). Gender bias in children’s perceptions of personality traits. Sex
Roles, 32, 17–28.
Powlishta, K., Serbin, L. A., Doyle, A., & White, D. R. (1994). Gender, ethnic,
and body type biases: The generality of prejudice in childhood. Developmental
Psychology, 30, 526–536.
Reicher, S. D., Spears, R., & Postmes, T. (1995). A social identity model of
de-individuation phenomena. European Review of Social Psychology, 6, 161–198.
Rutland, A. (1999). The development of national prejudice, in-group favoritism and
self stereotypes in British children. British Journal of Social Psychology, 38, 55–70.
Rutland, A., & Brown, R. (2001). Stereotypes as justifications for prior intergroup
discrimination: Studies of Scottish national stereotyping. European Journal of
Social Psychology, 31, 127–141.
Rutland, A., Cameron, L., Milne, A., & McGeorge, P. (2002, June). The development
and regulation of children’s implicit and explicit intergroup attitudes. Paper presented
at the 13th General Meeting of the European Association of Experimental Social
Psychology (EAESP), San Sebastián, Spain.
 Rutland 265
Rutland, A., Morrison, E., & Arnold-Dorman, C. (2001). The development of implicit
and explicit weight bias in girls and young females. Unpublished manuscript,
University of Kent, Canterbury, UK.
Scheier, M. F., & Carver, C. S. (1981). Private and public aspects of self. In L. Wheeler
(Ed.), Review of personality and social psychology, Vol. 2. London: SAGE
Publications.
Serbin, L. A., Conner, J. M., Burchardt, C. J., & Citron, C. C. (1979). Effects of peer
presence on sex-typing of children’s play behaviour. Journal of Experimental Child
Psychology, 27, 303–309.
Spears, R. (2001). The interaction between the individual and the collective self:
Self-categorization in context. In C. Sedikides & M. B. Brewer (Eds.), Individual
self, relational self, and collective self: Partners, opponents or strangers? Philadelphia,
PA: Psychology Press.
Spielman, D. A. (2000). Young children, minimal groups, and dichotomous
categorization. Personality and Social Psychology Bulletin, 26, 1433–1441.
Tajfel, H., Jahoda, G., Nemeth, C., Rim, Y., & Johnson, N. (1972). The devaluation by
children of their own national and ethnic group: Two case studies. British Journal of
Social and Clinical Psychology, 11, 235–243.
Tajfel, H., Nemeth, C., Jahoda, G., Campbell, J. D., & Johnson, N. (1970). The
development of children’s preference for their own country: A cross-national study.
International Journal of Psychology, 5, 245–253.
Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In W. G.
Austin & S. Worchel (Eds.), The social psychology of intergroup relations. Monterey,
CA: Brooks/Cole.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In
S. Worchel & W. G. Austin (Eds.), The psychology of intergroup relations (pp. 7–24).
Chicago, IL: Nelson-Hall.
Vaughan, G., Tajfel, H., & Williams, J. A. (1981). Bias in reward allocation in an
intergroup and an interpersonal context. Social Psychology Quarterly, 44, 37–42.
Verkuyten, M. (2001). National identification and intergroup evaluations in Dutch
children. British Journal of Developmental Psychology, 19, 559–571.
Wicklund, R. A., & Duval, S. (1971). Opinion change and performance facilitation as
a result of objective self-awareness. Journal of Experimental Social Psychology, 7,
319–342.
Yee, M., & Brown, R. J. (1994). The development of gender differentiation in young
children. Child Development, 33, 183–196.
10 Reducing stepfamily conflict:
The importance of inclusive
social identity
Brenda S. Banker, Samuel L. Gaertner,
John F. Dovidio, Missy Houlette,
Kelly M. Johnson, and Blake M. Riek

He drew a circle that shut me out –

Heretic, rebel, a thing to flout.
But love and I had the wit to win –
We drew a circle that took him in.
Edwin Markham

This chapter explores what happens when members of stepfamilies are asked,
as in Markham’s poem, to draw their circle of social inclusion wider. Does
love and harmony characterise the relations among those who are included?
We examine the consequences of this metaphor within a domain of group life
that provides a new, fundamental social identity for many people, the step-
family. Developing a stepfamily identity implicitly requires the expansion of a
person’s most basic and intimate social boundary, the biological family unit,
to include others, who arrive with their own sense of family identity. The
present chapter examines the role of group identity in establishing harmony
in stepfamilies and considers the factors that may inhibit or promote a more
inclusive family identity.
The stepfamily research we feature in this chapter was undertaken primar-
ily to explore the utility of the common in-group identity model (Gaertner &
Dovidio, 2000; Gaertner, Dovidio, Anastasio, Bachman, & Rust, 1993) for
increasing harmony between social groups. This model, derived from the
social categorisation approach (Allport, 1954; Tajfel, 1969), proposes that
factors that induce members of two groups to conceive of themselves as
members of a common, more inclusive in-group reduce intergroup conflict by
enabling cognitive and motivational processes that contribute to pro-in-group
favouring biases to be redirected to include former out-group members
(Gaertner, Mann, Murrell, & Dovidio, 1989). Before presenting the details of
our own stepfamily research, we first present our rationale for viewing the
stepfamily as an intergroup context and then we review the theoretical and
empirical evidence of the role of social categorisation in initiating intergroup
bias and its potential for producing intergroup harmony.
The logic of viewing the stepfamily as an intergroup context is derived
268 The importance of inclusive social identity
primarily from Allport (1954), who stated that the “biological family
ordinarily constitutes the smallest and firmest of one’s in-groups” (p. 41) and
that “every society on earth” regards the child “as a member of his parents’
groups” (p. 30). Members of first-married families generally share family
memories, ancestral histories and traditions, daily rituals, and a common
name, all of which contribute to a strong sense of family group identity
(e.g., see Settles, 1993). When divorce or death separates the first-married
couple, and remarriage to a new partner occurs, the biologically related par-
ents and their children who come together as a result of that remarriage do so
as two separate “in-groups”, with little or no common ground to bind them.
For many stepfamilies the unfortunate reality is that they remain fragments
of two separate families living together in conflict (Anthony, 1974).
Combining two groups into a single structure is complex and potentially
discordant. Whereas the initial development of each participating group
normally involved the interaction and socialisation of individuals who arrived
separately, stepfamily unions primarily represent a collision between groups.
Such mergers (see Gaertner, Bachman, Dovidio, & Banker, 2001), therefore,
are primarily an intergroup phenomenon. Moreover, families are particularly
influential groups in their members’ lives. They are fundamentally linked to
one’s social, economic, and psychological well-being and to personal and
collective identity.
Although interpersonal relations are not always smooth and harmonious,
intergroup relations are usually even more conflictual. Group members who
are interdependent with other groups are more distrusting and less likely to
be concerned with maximising joint outcomes compared to individuals who
share interdependence (Insko et al., 2001). As a consequence, the potential
for conflict and failure in stepfamilies is likely to be higher than in the forma-
tion of each original family (Haunschild, Moreland, & Murrell, 1994). In
addition, conflict between groups in stepfamilies is generally more intractable
than conflict between individuals because disputants arrive with a network of
social support for their respective positions. Ironically, in an atmosphere of
distrust and propensity for conflict, members of stepfamilies are expected to
identify with the new entity and to become committed to its well-being – with
the hopes of possibly living together “happily ever after”.
Unfortunately, stepfamily marriages have an extraordinarily high failure
rate. For second marriages the divorce rate is estimated at 60% (Norton &
Miller, 1992; White & Booth, 1985). This rate is even higher when children are
involved (Furstenberg & Spanier, 1984). Apparently, something very fre-
quently goes awry that threatens the success and longevity of the stepfamily.
Given the substantial emotional consequences that result from such failures,
it is important to understand some of the fundamental causes of these inter-
group catastrophes as well as factors that are related to successful marriages
between formerly separate family units. Indeed, there are reasons to believe
that the intergroup nature of these relationships, in part, places the success of
the stepfamily in jeopardy and it is the purpose of this chapter to examine the
 Banker et al. 269
stepfamily entity from the perspective of intergroup theory and research. In
particular we focus on the implications of the common in-group identity
model (Gaertner & Dovidio, 2000; Gaertner et al., 1993), a theoretical
framework we developed to conceptualise factors that are instrumental to the
creation and reduction of intergroup bias and conflict.

THE ROLE OF SOCIAL CATEGORISATION AND

INTERGROUP RELATIONS

Social categorisation has critical implications for one’s attitudes towards

others. Attraction and prejudice are fundamentally related to social categor-
isation and to the perception of intergroup boundaries – boundaries that
define who is included in one’s own group (a “we”) and who is excluded
(a “they”). Upon social categorisation, people favour in-group members in
terms of evaluations, attributions, material resources, helping, and social
support. Thus, changing the nature of ingroup inclusion and exclusion can
have important consequences for interpersonal and intergroup relations.
Social identity theory (SIT; Tajfel & Turner, 1979) and, more recently, self-
categorisation theory (SCT; Turner, 1985; see also Onorato & Turner, 2001)
consider the crucial role of social categorisation and identity in social
relations. With respect to one’s sense of identity, SIT and SCT view the
distinction between personal identity and social identity as a critical one (see
Spears, 2001). When personal identity is salient, a person’s individual needs,
standards, beliefs, and motives primarily determine behaviour. In contrast,
when social identity is salient, “people come to perceive themselves as more
interchangeable exemplars of a social category than as unique personalities
defined by their individual differences from others” (Turner, Hogg, Oakes,
Reicher, & Wetherell, 1987, p. 50). Under these conditions, collective needs,
goals, and standards are primary. Moreover, Tajfel and Turner proposed
that a person’s need for positive self-identity partially may be satisfied by
membership in prestigious social groups. Thus, this need motivates social
comparisons that favourably differentiate in-group from out-group members.
From a social categorisation perspective, when people or objects are cat-
egorised into groups, actual differences between members of the same cat-
egory tend to be perceptually minimised (Tajfel, 1969) and often ignored in
making decisions or forming impressions. Members of the same category
seem to be more similar than they actually are, and more similar than they
were before they were categorised together. In addition, between-group dif-
ferences tend to become exaggerated and overgeneralised, emphasising social
difference and group distinctiveness. The magnitude of these biases is a func-
tion of the salience of social categorisation (Abrams, 1985; Turner, 1985).
Moreover, these within- and between-group perceptual distortions have a
tendency to generalise to additional dimensions (e.g., character traits) beyond
those that originally differentiated the categories (Allport, 1954).
270 The importance of inclusive social identity
Whether people perceive others as a member of their group or of another
group impacts the emotional significance of group differences, and thus leads
to further perceptual distortion and to evaluative biases that reflect favour-
ably on the in-group and consequently on the self (Tajfel & Turner, 1979).
Upon social categorisation of individuals into in-groups and out-groups,
people spontaneously experience more positive affect towards other members
of the in-group (Otten & Moskowitz, 2000), particularly towards those who
are most prototypical of the group (Hogg & Hains, 1996). They also favour
in-group members directly in terms of evaluations and resource allocations
(Mullen, Brown, & Smith, 1992). In terms of cognitive processing, people
retain more information in a more detailed fashion for in-group members
than for out-group members (Park & Rothbart, 1982), have better memory
for information about ways in which in-group members are similar to and
out-group members are dissimilar to the self (Wilder, 1981), and remember
less positive information about out-group members (Howard & Rothbart,
1980). Perhaps due to the greater self–other overlap in representations for
people defined as in-group members (Smith & Henry, 1996), people process
information about and make attributions to in-group members more on the
basis of self-congruency than they do for out-group members (Gramzow,
L. Gaertner, & Sedikides, 2001).
Social categorisation also plays an important role in interpersonal and
intergroup attributions. People are more generous and forgiving in their
explanations for the behaviours of in-group relative to out-group members.
Positive behaviours and successful outcomes are more likely to be attributed
to internal, stable characteristics (the personality) of in-group than out-group
members, whereas negative outcomes are more likely to be ascribed to the
personalities of out-group members than of in-group members (Hewstone,
1990; Pettigrew, 1979). As a consequence, negative behaviours of out-group
members are less affected by mitigating evidence than are those of in-group
members (Beal, Ruscher, & Schnake, 2001). Relatedly, observed behaviours
of in-group and out-group members are encoded in memory at different
levels of abstraction (Maass, Salvi, Arcuri, & Semin, 1989). Undesirable
actions of out-group members are encoded at more abstract levels that pre-
sume intentionality and dispositional origin (e.g., she is hostile) than are
identical behaviours of in-group members (e.g., she slapped the girl). Desir-
able actions of out-group members, however, are encoded at more concrete
levels (e.g., she walked across the street holding the old man’s hand) relative
to the same behaviours of in-group members (e.g., she is helpful).
These cognitive biases help to perpetuate social biases and stereotypes
even in the face of countervailing evidence. For example, because positive
behaviours of out-group members are encoded at relatively concrete levels, it
becomes less likely that counterstereotypic positive behaviours would gener-
alise across situations or other out-group members (see also Karpinski &
Von Hippel, 1996). People do not remember that an out-group member was
“helpful”, but only the very concrete descriptive actions. Thus, out-group
 Banker et al. 271
stereotypes containing information pertaining to traits, dispositions, or
intentions are not likely to be influenced by observing counterstereotypic
out-group behaviours.
Language plays another role in intergroup bias through associations with
collective pronouns. Collective pronouns such as “we” or “they”, which are
used to define people’s in-group or out-group status, are frequently paired
with stimuli having strong affective connotations. As a consequence, these
pronouns may acquire powerful evaluative properties of their own. These
words (we, they) can potentially increase the availability of positive or nega-
tive associations and thereby influence beliefs about, evaluations of, and
behaviours towards other people, often automatically and unconsciously
(Perdue, Dovidio, Gurtman, & Tyler, 1990).
The process of social categorisation, however, is not completely unalter-
able. Categories are hierarchically organised, with higher-level categories
(e.g., nations) being more inclusive of lower-level ones (e.g., cities or towns).
By modifying a perceiver’s goals, motives, perceptions of past experiences,
and expectations, as well as factors within the perceptual field and the
situational context more broadly, there is an opportunity to alter the level of
category inclusiveness that will be primary or most influential in a given
situation. This malleability of the level at which impressions are formed is
important because of its implications for altering the way people think about
members of in-groups and out-groups, and consequently about the nature of
intergroup relations. Because categorisation is a basic process that is funda-
mental to intergroup bias, social psychologists have targeted this process as a
starting point to begin to improve intergroup relations. In the next section we
explore how the forces of categorisation can be harnessed and redirected
towards the reduction, if not the elimination, of intergroup bias.

REDUCING INTERGROUP BIAS AND CONFLICT

For almost 60 years, the contact hypothesis (Allport, 1954; see also Pettigrew,
1998) has represented the primary strategy for reducing intergroup bias and
facilitating harmony between groups. This hypothesis proposes that for con-
tact between groups to reduce bias successfully, certain prerequisite features
must be present. These characteristics of contact include equal status
between the groups, cooperative (rather than competitive) intergroup inter-
action, opportunities for personal acquaintance between the members, espe-
cially with those whose personal characteristics do not support negative
stereotypic expectations, and supportive norms by authorities within and
outside of the contact situation (Pettigrew, 1998). Research in laboratory and
field settings generally supports the efficacy of this list of prerequisite condi-
tions for achieving improved intergroup relations (see Pettigrew & Tropp,
2000). In the stepfamily literature, as well, research indicates that cooperation
in the stepfamily, as well as positive interaction between the steprelatives, are
272 The importance of inclusive social identity
associated with better overall stepfamily functioning and happiness (e.g., see
Anderson & White, 1986; Crosbie-Burnett, 1984; James & Johnson, 1987).
Recent research, however, has moved beyond specifying what conditions
moderate the reduction of bias to understanding what underlying processes,
such as those involving social categorisation, may be involved (see Pettigrew,
1998). From the social categorisation perspective, the issue to be addressed is
how intergroup contact can be structured to alter inclusive–exclusive collect-
ive representations of others. Two of the approaches that have been proposed
involve decategorisation and recategorisation. Decategorisation refers to
influencing whether people identify themselves primarily as group members
or as distinct individuals on the continuum proposed by Tajfel and Turner
(1979; see also Brewer, 1988; Brewer & Miller, 1984; Fiske, Lin, & Neuberg,
1999). Recategorisation, in contrast, is not designed to reduce or eliminate
categorisation, but rather to structure a definition of group categorisation at
a higher level of category inclusiveness in ways that reduce intergroup bias
and conflict (Allport, 1954, p. 43; see Dovidio, Gaertner, & Kafati, 2000;
Gaertner & Dovidio, 2000).
In each case, reducing the salience of the original inclusive–exclusive group
boundaries is expected to decrease intergroup bias. With decategorisation,
group boundaries are degraded, inducing members of different groups to
conceive of themselves and others as separate individuals (Wilder, 1981) and
encouraging more personalised interactions. When personalised interactions
occur, people “attend to information that replaces category identity as the
most useful basis for classifying each other” (Brewer & Miller, 1984, p. 288),
and thus category-based biases are reduced.
With recategorisation as proposed by the common in-group identity model
(Gaertner & Dovidio, 2000; Gaertner et al., 1993), inducing members of
different groups to conceive of themselves as a single, more inclusive super-
ordinate group rather than as two completely separate groups produces
attitudes towards former out-group members that become more positive
through processes involving pro-in-group bias. That is, the processes that lead
to favouritism towards in-group members would now be directed towards
former out-group members as they become redefined from exclusive to
inclusive categories.
The decategorisation and recategorisation strategies and their respective
means of reducing bias were directly examined in a laboratory study
(Gaertner et al., 1989). In this experiment, members of two separate
laboratory-formed groups were induced through various structural interven-
tions (e.g., seating arrangement) either to decategorise themselves (i.e., con-
ceive of themselves as separate individuals) or to recategorise themselves as
one superordinate group. Supporting the proposed value of altering the level
of category inclusiveness, these changes in the perceptions of intergroup
boundaries reduced intergroup bias. Furthermore, as expected, these strat-
egies reduced bias in different ways. Decategorising members of the two
groups reduced bias by decreasing the attractiveness of former in-group
 Banker et al. 273
members. In contrast, recategorising in-group and out-group members as
members of a more inclusive group reduced bias by increasing the attractive-
ness of the former out-group members. Consistent with SCT, “the attractive-
ness of an individual is not constant, but varies with the ingroup membership”
(Turner, 1985, p. 60).
In the next section, we present support for the common in-group identity
model and the effects of recategorisation. In addition, we discuss the value of
a “dual identity” in which original group identities are maintained but within
the context of a superordinate identity.

The common in-group identity model

In the common in-group identity model, we outline specific potential ante-
cedents and outcomes of direct or symbolic intergroup contact, as well as
identify possible mediating social categorisation processes. In particular,
we hypothesise that the different types of intergroup interdependence and
cognitive, perceptual, linguistic, affective, and environmental factors can
either independently or in concert alter individuals’ cognitive representations
of the aggregate. These resulting cognitive representations (i.e., one group,
two subgroups with one group, two groups, or separate individuals) are then
proposed to produce specific cognitive, affective, and overt behavioural con-
sequences. Thus, the causal factors (which include features specified by the
contact hypothesis) are proposed to influence members’ cognitive representa-
tions of the memberships that in turn mediate the relationship, at least in
part, between the causal factors and the cognitive, affective, and behavioural
consequences. In addition, we propose that common in-group identity can be
achieved by increasing the salience of existing common superordinate mem-
berships (e.g., a school, a company, a nation) or by introducing factors (e.g.,
common goals or shared fate) that are perceived to be shared between the
original groups.
Once out-group members are perceived as in-group members, it is pro-
posed that they would be accorded the benefits of in-group status. There
would probably be more positive thoughts, feelings, and behaviours towards
these former out-group members by virtue of their recategorisation as in-
group members. These more favourable impressions of out-group members
are not likely to be finely differentiated, at least initially (see Mullen & Hu,
1989). Rather, we propose that these more elaborated, personalised impres-
sions can soon develop within the context of a common identity because the
newly formed positivity bias is likely to encourage more open communication
and greater self-disclosing interaction between former out-group members.
Thus, as proposed by Pettigrew (1998; see also Hewstone, 1996), the nature of
intergroup contact can influence group representations in different ways
sequentially across time. We suggest that, over time, a common identity can
lead to decategorisation and encourage personalisation of out-group mem-
bers and thereby initiate a second route to achieving reduced bias. That is,
274 The importance of inclusive social identity
when viewed over time, decategorisation, recategorisation, and mutual inter-
group differentiation that preserves group identities are strategies for
reducing intergroup bias that are not competitors. Rather they can each con-
tribute to the reduction of intergroup bias and they can also reciprocally
facilitate each other (see also Hewstone, 1996). Thus, recategorisation can
lead to more interpersonally friendly, self-revealing interactions (e.g., see
Dovidio, Gaertner, Validzic, Matoka, Johnson, & Frazier 1997) and, as Hew-
stone speculated, a “common ingroup identity can affect decategorisation
(and possibly, over time differentiation, too)” (p. 354).
Within this model, we also acknowledge that the development of a com-
mon in-group identity does not necessarily require each group to forsake its
less inclusive group identity completely. As Brewer (2000) noted, individuals
belong simultaneously to several groups and possess multiple potential iden-
tities. As depicted by the “sub-groups within one group” (i.e., a dual identity)
representation, we believe that it is possible for members to conceive of two
groups (for example, parents and children) as distinct units within the context
of a superordinate (i.e., family) identity. This aspect of the common in-group
identity model is compatible with, although not identical to, other models
that propose that maintaining the salience of intergroup distinctions can be
important for producing generalised and longer-term reductions in inter-
group bias (Hewstone & Brown, 1986). In the next section, we examine some
empirical tests of the common in-group identity model.
Among the antecedent factors proposed by the common in-group identity
model are the features of contact situations (Allport, 1954) that are necessary
for intergroup contact to be successful (e.g., interdependence between groups,
equal status, egalitarian norms). From our perspective, cooperative inter-
action may enhance positive evaluations of out-group members and reduce
intergroup bias, at least in part, by transforming interactants’ representations
of the memberships from two groups to one group.
In one test of this hypothesis, we conducted an experiment that brought
two three-person laboratory groups together under conditions designed to
vary independently (1) the members’ representations of the aggregate as one
group or two groups through manipulation of the contact situation, and (2)
the presence or absence of intergroup cooperative interaction (Gaertner,
Mann, Dovidio, Murrell, & Pomare, 1990). The interventions designed to
emphasise common group membership through structural changes in the
contact situation (e.g., integrated vs. segregated seating; a new group name
for all six participants vs. the original group names; the same or different
colored tee shirts for both groups) and to encourage cooperative interaction
(joint evaluation and reward vs. independent outcomes) both reduced inter-
group bias. Moreover, they did so through the same mechanism. Contextual
features emphasising common “groupness” and joint outcomes each
increased one-group representations (and reduced separate-group representa-
tions), which in turn related to more favourable attitudes towards original
out-group members and lower levels of bias. Consistent with the common
 Banker et al. 275
in-group identity model, more inclusive, one-group representations mediated
the relationship between the interventions and the reduction of bias.
The advantage of this experimental design is that interdependence pre-
ceded changes in participants’ representations of the aggregate from two
groups to one group and also changes in intergroup bias. In addition, because
the representations of the aggregate were manipulated in the absence of
interdependence, the development of a one-group representation preceded
changes in intergroup bias. Thus, we can be confident about the directions of
causality in this study. In other experiments we have manipulated structural
aspects of intergroup contact situations, such as segregated or integrated
seating, and found that these manipulations directly influenced group repre-
sentations and, ultimately, intergroup bias (see Gaertner & Dovidio, 2000).
These studies provide convergent evidence for a clear sequence of events and
identification of causal relations.
An additional laboratory study (Houlette & Gaertner, 1999) explored the
effects of varying the degree of inclusiveness of a superordinate identity on
intergroup bias between three-person groups of liberals and conservatives.
The recategorisation manipulation emphasised members’ common affiliation
with their ad hoc six-person laboratory work group, their identity as Uni-
versity of Delaware students, or their common national citizenship (i.e.,
America). Intergroup bias decreased directly with greater inclusiveness of
their common superordinate entity (i.e., from the six-person work group to
national citizenship). Because participants’ identification with these super-
ordinate entities was also stronger with increasing levels of inclusiveness,
we cannot be certain whether reduced bias was driven by increased group
identification or by the greater inclusiveness of the superordinate entity per
se. In either case, however, these findings provide additional experimental
evidence of the effectiveness of interventions designed to emphasise common
group membership for reducing intergroup bias.
In a subsequent series of studies we utilised survey techniques under more
naturalistic circumstances to examine the impact of common group identity
for a range of types of intergroup bias. These studies offer converging sup-
port for the hypothesis that the features specified by the contact hypothesis
reduce intergroup bias, in part, because they transform members’ representa-
tions of the memberships from separate groups to one more inclusive group.
Participants in these studies included students attending a multi-ethnic high
school (Gaertner, Rust, Dovidio, Bachman, & Anastasio, 1996), and banking
executives who had experienced a corporate merger involving a wide variety
of banks across America (Bachman, 1993).
To provide a conceptual replication of the laboratory study of cooper-
ation, the surveys included items (specifically designed for each context) to
measure participants’ perceptions of the conditions of contact (i.e., equal
status, self-revealing interaction, cooperation, and egalitarian norms), their
representations of the aggregate (i.e., one group, two subgroups within one
group, two separate groups, and separate individuals), and a measure of
276 The importance of inclusive social identity
intergroup bias (see Gaertner, Dovidio, Nier, Ward, & Banker, 1999). For the
high school study, the main outcome measure was a bias in affective reactions,
such as feeling good and respectful towards in-group as compared to out-
group members. For the corporate merger study, the main measure of bias
was the extent to which the corporate executives perceived members of the
different organisations to have varying levels of positive work-related trait
characteristics (i.e., intelligent, hardworking, organised, skilled, and creative).
Consistent with the role of an inclusive group representation that is
hypothesised in the common in-group identity model, across these studies (1)
conditions of intergroup contact that were perceived as more favourable pre-
dicted lower levels of intergroup bias, (2) more favourable conditions of con-
tact predicted more inclusive (one group) and less exclusive (different groups)
representations, and (3) more inclusive representations of mediated lower
levels of intergroup bias and conflict (see Gaertner et al., 1999).
While supportive of the basic framework of our model, these cross-
sectional survey studies in natural group settings do not establish the direc-
tion of causality among conditions of contact, group representations, and
reductions in bias. Longitudinal methods that investigate the relationships
among these variables over time provide more direct insight into the direction
of causality. In the next section, in two studies that we present in greater
detail, we examine these hypothesised relationships in the context of step-
families. One of these studies, using cross-sectional data as in our earlier
survey studies, employs causal modelling to test the predicted effects. The
other study extends this line of research, and further tests for the proposed
causal sequence between the constructs in the model by using a longitudinal
survey methodology.

STEPFAMILY MARRIAGES

As we noted at the beginning of this chapter, families involve a primary form

of group membership. Beyond genetic relatedness, there is a profound social
connection among family members. Biological, or first-married families
generally share memories, ancestral histories, traditions, daily rituals, and
a common family name that contribute to a strong sense of family group
identity (see Settles, 1993).
Given the importance of family group identity, stepfamilies represent an
interesting domain in which to examine the utility of the common in-group
identity model. When the first-married family is fragmented by divorce or
death and remarriage to a new partner occurs, the biological parents and
children from the two families come together under new circumstances.
Through remarriage, members of two separate “in-groups”, with no common
memories, histories, daily rituals, or even family name find themselves in an
intensive intergroup context.
Relative to biologically related, first-married families, stepfamilies have
 Banker et al. 277
generally been described as more stressful and less cohesive. Stepfathers,
for example, not only report being less satisfied with their own lives than do
first-married fathers, they also indicate that the lives of their stepchildren
are less than satisfactory as well (Fine, McKenry, Donnelly, & Voydanoff,
1992). Bray and Berger (1993) found that, in couples remarried for 5 to 7
years, there were less positive wife-to-husband and biological parent–child
interactions in stepfamilies than in their first-married counterparts. And
several studies have found that stepparent–stepchild relationships are more
conflict-ridden than are those between biological parents and children in
first-married families (Anderson & White, 1986; Furstenberg, 1987; Sauer &
Fine, 1988).
One reason why stepfamilies may experience greater conflict is less than
satisfactory contact among the stepfamily members. James and Johnson
(1987) reported that competitiveness in stepfamilies is related to marital dis-
satisfaction and psychological pathology in both husbands and wives.
Cooperativeness, however, relates to marital satisfaction for both partners,
and to the husbands’ positive psychological adjustment. Furthermore, the
failure of stepchildren to respond in kind to their stepparents’ positive
behaviours towards them has been found to be associated with stepfamily
dysfunction (Anderson & White, 1986; Brown, Green, & Druckman, 1990).
A more positive relationship between the stepparent and stepchild, in con-
trast, is associated with more positive stepfamily functioning (Anderson &
White, 1986) and happiness (Crosbie-Burnett, 1984).

Stepfamily relations
In our first study with stepfamilies (Banker & Gaertner, 1998), using a cross-
sectional, rather than a longitudinal, design, our goals were (a) to investigate
the validity of our vision of stepfamilies as an intergroup entity and (b) to
empirically support the utility of the common in-group identity model in this
intergroup context. Consistent with these goals, we hypothesised that (1)
stepfamily members would see their stepfamilies as being more like two sep-
arate families and less like one family than first-married families, (2) there
would be a significant relationship between positive conditions of contact in
the stepfamily household and increased stepfamily harmony, and (3) the rela-
tionship between positive conditions of contact and increased harmony
would be mediated by a one-family representation of the stepfamily such that
the greater the one-family representation, the greater the harmony.
To test these hypotheses, we surveyed university students who identified
themselves as either living at home in a “complex” stepfamily (i.e., the remar-
ried parents each have at least one child from their previous marriage so as to
create an “intergroup” household) or in a first-married family. There were 86
stepfamily participants and 65 participants from first-married families. In 46
stepfamily households, the stepsibling(s) lived in the same household as the
participant full-time. In 40 households there was a visitation arrangement;
278 The importance of inclusive social identity
the average visitation was 4 days/month. All of the participants from
first-married families had at least one biologically related sibling at home.
The participants completed identical surveys containing items designed to
tap their perceptions of the conditions of contact in the household using 1–7
scales. This measure included items representing perceptions of cooperation
(e.g., “There is generally a spirit of competition rather than cooperation in
the house” [reverse scored], or “It feels like my stepsiblings are always com-
peting with my siblings and/or me” [reverse scored]), equality (e.g., “I feel that
I have as much private space at home as do the other people living in the
house” or “There are different household rules for me and my siblings than
there are for my stepsiblings” [reverse scored]), and opportunities for person-
alisation (e.g., “Holiday outings or vacations are usually planned so that
everyone in the household is included”). Participants also, using 1–7 ratings,
completed measures of the cognitive representation of the family (e.g., one
family: “Living in my house it feels like one family”) and family harmony
(e.g., “I would characterise the environment at my house as ‘harmonious’ ”).
A comparison of the responses from these two family types reveals that
stepfamilies felt more like two separate families (M = 2.24) than did first-
married families (M = 1.53) and less like one family (M = 5.18) than first-
married families (M = 5.97). These results provide initial empirical support
for our assumption that stepfamilies may commonly represent a naturalistic
intergroup entity. In addition, the correlations among the variables were con-
sistent with the hypotheses derived from the common in-group identity
model. As expected, more favourable conditions of contact in the stepfamily
related to higher levels of stepfamily harmony (r = .54). More favourable
conditions of contact also related to stronger perceptions of the stepfamily as
one family (r = .75) and to weaker perceptions of the stepfamily as two
separate families (r = –.73).
Taken together, these results satisfy preliminary requirements of mediation
analysis1 (Baron & Kenny, 1986). That is, the independent variable (condi-

1 According to Baron and Kenny (1986), the idea of mediation originated with Woodworth’s
(1928, cited in Baron & Kenny, p. 1176) S-O-R model, which proposes that some psychological
transformation within the organism intervenes between a stimulus and a response. This inter-
vening variable represents the mechanism or process by which the stimulus (the independent
variable) is hypothesised to influence the response (the dependent variable). For example, in an
experiment discussed earlier in this chapter (Gaertner et al., 1990), we hypothesised that
cooperative intergroup interaction reduces intergroup bias because cooperation changes mem-
bers’ cognitive representations of the aggregate from two groups to one group. In that study,
the independent variable that we manipulated (the stimulus) involved whether the contact
between the two groups that we arranged represented cooperative or independent activities.
The dependent variable that we measured (the response) was intergroup bias. The proposed
mediating variable (or the intervening organismic variable in the S-O-R model) was members’
perceptions of the aggregate as one group. Statistically, Baron and Kenny outline a procedure
involving a series of multiple regression analyses to demonstrate mediation. The first regres-
sion tests whether the independent variable predicts the dependent variable. Mediation can
 Banker et al. 279
tions of contact) predicted the dependent variable (stepfamily harmony) and
the proposed mediating variable, more inclusive cognitive representations (e.g.,
one-family representation of the stepfamily). Next, using AMOS hardware
(Arbuckle, 1997), we directly tested the critical mediation hypothesis. Provid-
ing evidence of at least partial mediation, the path from conditions of contact
to stepfamily harmony was weaker (Beta = .37) when cognitive representations
were considered simultaneously as predictors in the regression equation.
Moreover, the indirect path from the conditions of contact through one family
to stepfamily harmony was significant (Beta = .19). Thus, following the pro-
posed sequence, the more positive were the conditions of contact, the more the
stepfamily felt like one family and the greater was the stepfamily harmony.
The results of this study (Banker & Gaertner, 1998) provide strong initial
support for the utility of our intergroup relations perspective in modelling the
reduction of stepfamily conflict. Because these data are cross-sectional, as we
noted earlier, the results are correlational: We cannot be certain about the
direction of causality of the variables. That is, while the results suggest that a
strong one-group cognitive representation of the stepfamily leads to lower
stepfamily conflict, it may very well be that lower stepfamily conflict makes
the stepfamily feel more like one family. Our next study, presented in the
following section, addresses this issue by using a longitudinal design.

Stepfamily relations over time

A primary focus of our longitudinal study was to examine more directly the
hypothesised direction of causality of the variables specified by the contact
hypothesis (Allport, 1954; Cook, 1985) and the common in-group identity
model (Gaertner & Dovidio, 2000) in stepfamilies. Over time, positive condi-
tions of contact were expected to reduce stepfamily conflict by making the
stepfamily feel more like one family than two. Again, we hypothesised that
the stronger the one-family representation, the less conflict there would be in
the stepfamily household.
For this study, we recruited members of complex stepfamilies through
appeals to classes, through newspaper advertisements, and through contacts
made with the Stepfamily Association of America. In the first phase of the
study (Time 1), there were 148 stepchild participants (53 male and 95
female). In the second wave (Time 2), conducted 9 months later, there were 66

only be statistically demonstrated if the independent variable (the stimulus) is first shown to
influence the dependent variable (the response) significantly. Next, to show mediation, the
second regression must reveal that the independent variable significantly predicts the proposed
mediating variable. Then, in a third regression equation in which the independent variable and
the mediating variable are considered simultaneously as predictors of the dependent variable,
for mediation to occur the mediating variable should emerge as a significant predictor while
the effect of the independent variable should be significantly weaker than in the first regression
equation, and with complete mediation, nonsignificant.
280 The importance of inclusive social identity
participants (44.6% of Phase 1; 30 male and 36 female). The average age of
the participants at Time 1 was 18.3 years; at Time 2, it was 18.7 years. At
Time 1, there were 60 households in which the stepsibling(s) lived full-time
and 88 households in which they visited. The average yearly visitation was 3.7
months. At Time 2, there were 26 households in which the stepsibling(s) lived
full-time and 40 households in which they visited. The average yearly
visitation was 4.0 months.
The surveys used in the two phases of this research were similar to those
used in our previous stepfamily study. These surveys included an 11-item
measure of conditions of contact (Cronbach alphas = .91 at each time),
cognitive representations of the stepfamily (one family: “Living in my house,
it feels like one family”; two separate families: “Living in my house, it feels
like two separate families.”), and a 4-item measure of stepfamily harmony/
conflict (e.g., “There is conflict between individuals from Family A and B in
the household”; Cronbach alphas = .77 and .82 at Time 1 and Time 2).
Consistent with the results of our earlier stepfamily study, more favourable
conditions of contact were related to higher ratings of one family (r = .68)
and to lower ratings of two separate families (r = –.58). More favourable
conditions of contact were also significantly correlated with less conflict (i.e.,
more harmony) in the household (r = –.69). Similar relations were obtained
among these measures at Time 2: rs = .67, –.52, and –.67.
On the basis of the data from the 66 participants who were studied across
the 9-month period, we tested the hypothesised causal relationships (using
AMOS software; Arbuckle & Wothke, 1999). Specifically, we examined the
stepchildren’s longitudinal panel data in a series of cross-lagged structural
equation models to assess the degree of empirical support there is for our
primary hypotheses. We expected that, over time, positive conditions of con-
tact at Time 1 would lead to a decrease in stepfamily conflict measured at
Time 2, and to a stronger one-family representation of the stepfamily at Time
2. This would be more supportive of our model than if conflict at Time 1
strongly predicted the conditions of contact at Time 2, although we recognise
that these relationships are probably bidirectional. Finally, we hypothesised
that a one-family representation of the stepfamily at Time 1 would lead to a
reduction in perceived stepfamily conflict at Time 2. These analyses
controlled for demographic variables, such as the respondent’s age2.

2 The structural components of the cross-lagged models include the covariance between the
Time 1 variables and four paths: two autoregressive paths from each Time 1 variable to its
Time 2 counterpart (e.g., Contact 1 to Contact 2), and the two cross-lagged paths from
each Time 1 variable to the other variable at Time 2 (e.g., Contact 1 to Conflict 2, and Conflict
1 to Contact 2). Because several demographic variables (e.g., child’s age) were found to signifi-
cantly relate to the variables of interest, each was tested as a covariate. These demographic
variables were entered into the structural equation model as covariates when appropriate. Each
of these cross-lagged models fit our data very well in terms of the chi-square and goodness of
fit indices (i.e., GFI, AGFI, CFI, and RMSEA).
 Banker et al. 281

Figure 10.1 Longitudinal analysis: Contact and conflict

As hypothesized, the results (see Figure 10.1) reveal that positive condi-
tions of contact at Time 1 significantly predict a decrease in conflict (Beta =
–.234) at Time 2, controlling for initial levels of conflict in the household at
Time 1. Also, in terms of the direction of causality being reversed from what
we expected, conflict at Time 1 did not significantly predict conditions of
contact at Time 2 (Beta = –.13). Also, in support of the direction of causality
specified by the common in-group identity model, we hypothesised that posi-
tive conditions of contact at Time 1 would lead to increased perceptions of
the stepfamily as one family at Time 2. The results of the cross-lagged path
analysis (see Figure 10.2) reveal that, as hypothesised, positive conditions of
contact at Time 1 reliably related to an increase in the one-family representa-
tion of the stepfamily at Time 2 (Beta = .254), controlling for the level of the
one-family representation at Time 1. Also, the nonpredicted path from one-
family representation at Time 1 to conditions of contact at Time 2 was not
significant.
The idea that is central to the common in-group identity model is the
expectation that, when two groups interact, the relationship between the con-
tact conditions and intergroup conflict is mediated by the cognitive represen-
tation of the groups: The more the groups feel like “one group”, the less the
intergroup conflict. To test this mediation hypothesis longitudinally, we pre-
dicted that a one-family representation of the stepfamily at Time 1 would
significantly relate to decreased conflict at Time 2. The results of the cross-
lagged structural equation analysis (see Figure 10.3) reveals a significant path
indicating, as predicted, that the more the stepfamily feels like one family at
282 The importance of inclusive social identity

Figure 10.2 Longitudinal analysis: Contact and one family

Figure 10.3 Longitudinal analysis: One family and conflict

 Banker et al. 283
Time 1, the less conflict there is at Time 2 (Beta = –.305), controlling for the
level of conflict at Time 1. Furthermore, the nonpredicted path from conflict
at Time 1 to one family at Time 2 was not significant, and additional analyses
strongly suggested that the path from one family at Time 1 to conflict at Time
2 was unidirectional. Therefore, extrapolating from the causal paths
described in the above analyses (see Figures 10.1, 10.2, and 10.3), there is
general support for the idea proposed by the common in-group identity
model that over time, the conditions of contact lead to changes in the cogni-
tive representation of the stepfamily as one family, and that the representa-
tion of the stepfamily as one family leads to a reduction in stepfamily
conflict.

CONCLUSIONS

In our research with stepfamilies, we utilise the perspective of the common

in-group identity model (Gaertner & Dovidio, 2000) to investigate and model
the pursuit of stepfamily harmony. Our results (Banker & Gaertner, 1998)
indicate that stepfamilies feel more like two separate families and less like one
family than do first-married families, and that this is problematic for step-
family members. These earlier findings reveal, for example, that positive con-
ditions of contact in the stepfamily household relate to feeling more like
one family, which in turn relates to greater harmony in the stepfamily
household. Together, the results of Banker and Gaertner and our recent
longitudinal research suggest that viewing one’s stepfamily as “one family”
may be an important causal link in the attainment of stepfamily har-
mony and the reduction of stepfamily conflict. Indeed, results of our
longitudinal study indicate that having a one-family representation of the
stepfamily at Time 1 significantly relates to a reduction in stepfamily conflict
at Time 2.
The results of our longitudinal study support the proposed direction of
causality of the variables specified by the common in-group identity model
(Gaertner & Dovidio, 2000) in the naturally occurring intergroup context of
the stepfamily. Specifically, positive conditions of contact in the stepfamily
household at Time 1 lead to less conflict at Time 2 and a one-family represen-
tation of the stepfamily at Time 2. Also, having a one-family representation
of the stepfamily at Time 1 leads to less stepfamily conflict at Time 2.
There remains, however, a possibility that for these stepchildren the causal
systems under investigation actually involve reciprocal feedback over time,
for example, between conditions of contact and stepfamily conflict. That is,
in reality, favourable conditions of contact reduce stepfamily conflict, and the
decrease in conflict may, in turn, strengthen the favourableness of the condi-
tions of contact. The potential of such feedback systems is quite plausible
within or outside of the consideration of our model. In future studies,
three waves of data collection may be necessary to verify or negate these
284 The importance of inclusive social identity
suppositions. Nevertheless, with two waves of data available, Finkel (1995)
suggests that “when a causal system is one of continuous reciprocal feed-
back” . . . “the cross-lagged model ‘tends not to be misleading about the
direction of causal influence’ ” (p. 25). In any event, the results of our longi-
tudinal analyses suggest that helping stepfamily members to see their step-
family as one family, or as one common in-group, can only help in the
stepfamily’s pursuit of harmony. Apparently, factors that promote step-
family harmony, or decrease stepfamily conflict, in part cast wider circles of
inclusion.
The results of our research offer support for the common in-group identity
model and its utility for studying stepfamily marriages. Moreover, the find-
ings support and extend previous work in family research and suggest ways
that stepfamily harmony can be achieved. Indeed, the mediation by the
one-family representation of the stepfamily of the relationships between
the conditions of contact and harmony in the stepfamily, and between the
conditions of contact and stepfamily conflict, suggest that the process of
recategorisation from two separate groups to one, single, inclusive family is
central to harmonious stepfamily relations. Supportive of the model, the
more the stepfamily feels like one family, the less the conflict and the greater
the harmony in the stepfamily.
These results also offer an explanation for how the lack of cooperative
interaction among stepfamily members (James & Johnson, 1987) may
become associated with stepfamily dysfunction or disharmony. In the future,
other stepfamily issues (e.g., family economics) may also be identified as
possibly influencing stepfamily members’ perceptions of the stepfamily unit
and harmony. For example, it was suggested by Fishman (1983) that step-
families who pool their financial resources for the stepfamily’s general use
regardless of biological ties (i.e., the “common pot economy”) are more uni-
fied than are those who maintain separate bank accounts (i.e., the “two pot
economy”) that are used for one’s own and one’s biological relatives’ needs
(p. 359). From the perspective of the common in-group identity model,
Fishman’s common pot economy would be expected to contribute to the
perception of the stepfamily as one family and to yield greater stepfamily
harmony. The identification of activities and relationships that contribute to
the perception of the stepfamily as one family also might one day aid in the
clinical evaluation of stepfamily functioning, as well as form the basis for new
therapeutic tools with which to understand and treat problems encountered
by stepfamily members. Likewise, the identification of factors that contribute
to the perception of the stepfamily as one family may also inform theories of
intergroup relations and provide new tools with which to understand and
heal intergroup problems across a variety of settings.
In conclusion, bringing families together to create a successful stepfamily is
a complex task with a substantial risk of failure. Because families are primary
groups closely connected to one’s social identity, successful stepfamily out-
comes require managing intergroup as well as interpersonal processes in
 Banker et al. 285
constructive and productive ways. Our research indicates that the develop-
ment of a common in-group identity – that is, changing the stepfamily
members’ perceptions of the stepfamily unit from “us” and “them” to a more
inclusive “we” – plays a central role in reducing stepfamily conflict and in
facilitating harmonious stepfamily relations.

Acknowledgements
Preparation of this chapter was supported by NIMH Grant MH 48721.

REFERENCES

Abrams, D. (1985). Focus of attention in minimal intergroup discrimination. British

Journal of Social Psychology, 24, 65–74.
Allport, G. W. (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley.
Anderson, J. Z., & White, G. D. (1986). An empirical investigation of interaction
and relationship patterns in functional and dysfunctional nuclear families and
stepfamilies. Family Process, 25, 407–422.
Anthony, E. J. (1974). Children at risk from divorce: A review. In E. J. Anthony &
C. Koupernik (Eds.), The child in his family, Vol. 3. New York: John Wiley & Sons.
Arbuckle, J. L. (1997). AMOS users’ guide: Version 3.6. Chicago: SPSS.
Arbuckle, J. L., & Wothke, W. (1999). AMOS users’ guide. Version 4.0. Chicago: SPSS.
Bachman, B. A. (1993). An intergroup model of organizational mergers. Unpublished
PhD dissertation, University of Delaware, Newark, DE.
Banker, B. S., & Gaertner, S. L. (1998). Achieving stepfamily harmony: An
intergroup-relations approach. Journal of Family Psychology, 12, 310–325.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51, 1173–1182.
Beal, D. J., Ruscher, J. B., & Schnake, S. B. (2001). No benefit of the doubt: Intergroup
bias in understanding causal explanation. British Journal of Social Psychology, 40,
531–543.
Bray, J. H., & Berger, S. H. (1993). Developmental issues in stepfamilies research
project: Family relationships and parent–child interactions. Journal of Family
Psychology, 7, 76–90.
Brewer, M. B. (1988). A dual process model of impression formation. In T. K. Srull &
R. S. Wyer, Jr. (Eds.), Advances in social cognition (Vol. 1, pp. 1–36). Hillsdale,
NJ: Lawrence Erlbaum Associates, Inc.
Brewer, M. B. (2000). Reducing prejudice through cross-categorization: The effects of
multiple social identities. In S. Oskamp (Ed.), Reducing prejudice and discrimination
(pp. 165–183). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Brewer, M. B., & Miller, N. (1984). Beyond the contact hypothesis: Theoretical per-
spectives on desegregation. In N. Miller & M. B. Brewer (Eds.), Groups in contact:
The psychology of desegregation (pp. 281–302). Orlando, FL: Academic Press.
Brown, A. C., Green, R. J., & Druckman, J. (1990). A comparison of stepfamilies with
and without child-focused problems. American Orthopsychiatric Association, 60,
556–566.
286 The importance of inclusive social identity
Cook, S. W. (1985). Experimenting on social issues: The case of school desegregation.
American Psychologist, 40, 452–460.
Crosbie-Burnett, M. (1984). The centrality of the step-relationship: A challenge to
family theory and practice. Family Relations, 33, 459–463.
Dovidio, J. F., Gaertner, S. L., & Kafati, G. (2000). Group identity and intergroup
relations: The common in-group identity model. In S. R. Thye, E. J. Lawler, M. W.
Macy, & H. A. Walker (Eds.), Advances in group processes, Vol. 17 (pp. 1–34).
Stamford, CT: JAI Press.
Dovidio, J. F., Gaertner, S. L., Validzic, A., Matoka, K., Johnson, B., & Frazier, S.
(1997). Extending the benefits of re-categorization: Evaluations, self-disclosure and
helping. Journal of Experimental Social Psychology, 33, 401–420.
Fine, M. A., McKenry, P. C., Donnelly, B. W., & Voydanoff, P. (1992). Perceived
adjustment of parents and children: Variations by family structure, race, and
gender. Journal of Marriage and the Family, 54, 118–127.
Finkel, S. E. (1995). Causal analysis with panel data. Thousand Oaks, CA: SAGE
Publications.
Fishman, B. (1983). The economic behavior of stepfamilies. Family Relations, 32,
359–366.
Fiske, S. T., Lin, M., & Neuberg, S. L. (1999). The continuum model: Ten years
later. In S. Chaiken & Y. Trope (Eds.), Dual process theories in social psychology
(pp. 231–254). New York: Guilford Press.
Furstenberg, F. F. Jr. (1987). The new extended family: The experience of parents and
children after remarriage. In K. Pasley & M. Ihinger-Tallman (Eds.), Remarriage
and stepparenting: Current research and theory (pp. 42–61). New York: Guilford
Press.
Furstenberg, F. F. Jr., & Spanier, G. B. (1984). Recycling the family: Remarriage after
divorce. Beverly Hills, CA: SAGE Publications..
Gaertner, S. L., & Dovidio, J. F. (2000). Reducing intergroup bias: The common
ingroup identity model. Philadelphia, PA: Psychology Press.
Gaertner, S. L., Bachman, B. A., Dovidio, J. D., & Banker, B. S. (2001). Corporate
mergers and stepfamily marriages: Identity, harmony, and commitment. In
M. A. Hogg & D. Terry (Eds.), Social identity processes in organization contexts
(pp. 265–282). Philadelphia, PA: Psychology Press.
Gaertner, S. L., Dovidio, J. F., Anastasio, P. A., Bachman, B. A., & Rust, M. C. (1993).
The common ingroup identity model: Recategorization and the reduction of
intergroup bias. In W. Stroebe & M. Hewstone (Eds.), European review of social
psychology, Vol. 4 (pp. 1–26). New York: John Wiley & Sons.
Gaertner, S. L., Dovidio, J. F., Nier, J., Ward, C., & Banker, B. (1999). Across cultural
divides: The value of a superordinate identity. In D. A. Prentice & D. T. Miller
(Eds.), Cultural divides: Understanding and overcoming group conflict (pp. 173–212).
New York: Russell Sage Foundation.
Gaertner, S. L., Mann, J. A., Dovidio, J. F., Murrell, A. J., & Pomare, M. (1990).
How does cooperation reduce intergroup bias? Journal of Personality and Social
Psychology, 59, 692–704.
Gaertner, S. L., Mann, J. A., Murrell, A. J., & Dovidio, J. F. (1989). Reduction of
intergroup bias: The benefits of recategorization. Journal of Personality and Social
Psychology, 57, 239–249.
Gaertner, S. L., Rust, M. C., Dovidio, J. F., Bachman, B. A., & Anastasio, P. A. (1996).
The contact hypothesis: The role of a common ingroup identity on reducing inter-
 Banker et al. 287
group bias among majority and minority group members. In J. L. Nye & A. M.
Brower (Eds.), What’s social about social cognition? (pp. 230–360). Newbury Park,
CA: SAGE Publications.
Gramzow, R. H., Gaertner, L., & Sedikides, C. (2001). Memory for in-group and out-
group information in a minimal group context: The self as an informational base.
Journal of Personality and Social Psychology, 80, 188–205.
Haunschild, P. R. , Moreland, R. L., & Murrell, A. J. (1994). Sources of resistance to
mergers between groups. Journal of Applied Social Psychology, 24, 1150–1178.
Hewstone, M. (1996). Contact and categorization: Social psychological interventions
to change intergroup relations. In N. Macrae, M. Hewstone, & C. Stangor (Eds.),
Foundations of stereotypes and stereotyping (pp. 323–368). New York: Guilford
Press.
Hewstone, M., & Brown, R. J. (1986). Contact is not enough: An intergroup perspec-
tive on the contact hypothesis. In M. Hewstone & R. Brown (Eds.), Contact and
conflict in intergroup encounters (pp. 1–44). Oxford: Basil Blackwell.
Hogg, M. A., & Hains, S. C. (1996). Intergroup relations and group solidarity: Effects
of group identification and social beliefs on depersonalized attraction. Journal of
Personality and Social Psychology, 70, 295–309.
Houlette, M., & Gaertner, S. L. (1999). The common ingroup identity model: Effects of
inclusiveness of common ingroup identity and subgroup salience on ingroup bias.
Unpublished data, Department of Psychology, University of Delaware, Newark,
DE.
Howard, J. M., & Rothbart, M. (1980). Social categorization for in-group and
out-group behavior. Journal of Personality and Social Psychology, 38, 301–310.
Insko, C. A., Schopler, J., Gaertner, L., Wildschut, T., Kozar, R., Pinter, B., Finkel,
E. J., Brazil, D. M., Cecil, C. L., & Montoya, M. R. (2001). Interindividual–
intergroup discontinuity reduction through the anticipation of future interaction.
Journal of Personality and Social Psychology, 80, 95–111.
James, S. D., & Johnson, D. W. (1987). Social interdependence, psychological adjust-
ment, and marital satisfaction in second marriages. Journal of Social Psychology,
128, 287–303.
Karpinski, A., & Von Hippel, W. (1996). The role of linguistic intergroup bias in
expectancy maintenance. Social Cognition, 4, 141–163.
Kessler, T., & Mummendey, A. (2001). Is there any scapegoat around? Determinants
of intergroup conflicts at different categorization levels. Journal of Personality and
Social Psychology, 81, 1102–1118.
Maass, A., Salvi, D., Arcuri, L., & Semin, G. R. (1989). Language use in inter-
group contexts: The linguistic intergroup bias. Journal of Personality and Social
Psychology, 57, 981–993.
Mullen, B., Brown, R. J., & Smith, C. (1992). Ingroup bias as a function of salience,
relevance, and status: An integration. European Journal of Social Psychology, 22,
103–122.
Mullen, B., & Hu, L. T. (1989). Perceptions of ingroup and outgroup variability: A
meta-analytic integration. Basic and Applied Social Psychology, 10, 233–252.
Norton, A. J., & Miller, L. F. (1992). Marriage, divorce and remarriage in the 1990s
(US Bureau of the Census, Current Population Reports. Series P-23, No. 180.)
Washington, DC: US Government Printing Office.
Onorato, R. S., & Turner, J. C. (2001). The “I”, “me”, and the “us”: The psychological
group and self-concept maintenance and change. In C. Sedikides & M. B. Brewer
288 The importance of inclusive social identity
(Eds.), Individual self, relational self, collective self (pp. 147–170). Philadelphia, PA:
Psychology Press.
Otten, S., & Moskowitz, G. B. (2000). Evidence for implicit evaluative in-group bias:
Affect-based spontaneous trait inference in a minimal group paradigm. Journal of
Experimental Social Psychology, 36, 77–89.
Park, B., & Rothbart, M. (1982). Perception of out-group homogeneity and levels of
social categorization: Memory for the subordinate attributes of in-group and
out-group members. Journal of Personality and Social Psychology, 42, 1051–1068.
Perdue, C. W., Dovidio, J. F., Gurtman, M. B., & Tyler, R. B. (1990). “Us”
and “them”: Social categorization and the process of intergroup bias, Journal of
Personality and Social Psychology, 59, 475–486.
Pettigrew, T. F. (1979). The ultimate attribution error: Extending Allport’s cognitive
analysis of prejudice. Personality and Social Psychology Bulletin, 55, 461–476.
Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49,
65–85.
Pettigrew, T. F., & Tropp, L. R. (2000). Does intergroup contact reduce prejudice?
Recent meta-analytic findings. In S. Oskamp (Ed.), Reducing prejudice and
discrimination (pp. 93–114). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Rugosa, D. (1988). Myths about longitudinal research. In K. W. Schaie, R. T.
Campbell, W. Meredith, & S. C. Rawlings (Eds.), Methodological issues in aging
research (pp. 171–209). New York: Springer.
Sauer, L. E., & Fine, M. A. (1988). Parent–child relationships in stepparent families.
Journal of Family Psychology, 1, 434–451.
Settles, B. H. (1993). The illusion of stability in family life: The reality of change and
mobility. Marriage and Family Review, 19, 5–29.
Smith, E. R., & Henry, S. (1996). An in-group becomes part of the self: Response time
evidence. Personality and Social Psychology Bulletin, 22, 635–642.
Spears, R. (2001). The interaction between the individual and the collective self:
Self-categorization in context. In C. Sedikides & M. B. Brewer (Eds.), Individual
self, relational self, collective self (pp. 171–198). Philadelphia, PA: Psychology Press.
Tajfel, H. (1969). Cognitive aspects of prejudice. Journal of Social Issues, 25, 79–97.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In
W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations
(pp. 33–48). Monterey, CA: Brooks/Cole.
Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive
theory of group behavior. In E. J. Lawler (Ed.), Advances ingroup processes, Vol. 2
(pp. 77–122). Greenwich, CT: JAI Press.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorization theory. Oxford: Basil Blackwell.
White, L. K., & Booth, A. (1985). The quality and stability of remarriages: The role of
stepchildren. American Sociological Review, 689–698.
Wilder, D. A. (1981). Perceiving persons as a group: Categorization and intergroup
relations. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup
behavior (pp. 213–257). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Part IV

Epilogue
11 The development of social
identity: What develops?
Dominic Abrams

The chapters in this book bring together an impressive body of research to

make an exciting and scholarly contribution to our understanding of the
development of social identity. They provide a clear signpost for questions to
be tackled in future research. As a whole, the book is an extremely valuable
resource for anyone wishing to research social identity as a developmental
phenomenon. Various themes emerge strongly, and some others find voice
but with less certainty and stridency. My aims in this commentary are to pose
three broad questions at the outset, consider each chapter in turn (devoting
most space to a discussion of Ruble et al., simply because they raise many of
the issues common to several chapters), then conclude by reflecting on how
effectively the broader questions have been explored.
The majority of chapters in the book focus on the cognitive and motiv-
ational elements of self-categorisation and social identity. First perhaps, these
should not be the only, or the most important, aspects of children’s group
membership that we need to consider. Are we asking the right questions
about children’s relationship to social groups and categories?
Second, the orientation to social identity development offered by chapters
in this book seems to focus more strongly either on social context or on
cognitive development. There might be other elements in the develop-
mental equation that are being neglected. Are we analysing social identity
comprehensively as a developmental phenomenon?
Third, there are controversies in the adult literature surrounding various
assumptions made by social identity theory and self-categorisation theory.
How far does the developmental research reflect these controversies?

THE CHAPTERS

Ruble et al.
Ruble, Alvarez, Bachman, Cameron, Fuligni, Garcia Coll, and Rhee
provide an excellent resource for anyone interested in social categorisation
in childhood, offering a careful analysis of the different ways in which
self-conception and intergroup behaviour are connected. (Nonetheless, there
292 The development of social identity
are several interesting points that could be added or developed further – see
below.) After referring to a variety of approaches to the social self, Ruble et
al. propose that it is collective identity that is of most interest in connection
with the development of categorisation in the cognitive-development
literature. They state that, “the nature of such categorisation and com-
parison processes are likely to change dramatically between the early and
mid elementary school years”.
It is an empirical question as to quite how dramatic the changes might be,
but there is also the question of why these changes might happen. For
example, a self-categorisation theory (SCT) line might be that the content and
focus of categorisation might change, and this might provide opportunities
for different social comparisons. However, once children are able to categor-
ise things, the fundamental cognitive process should not change (see also
David et al., Chapter 5). If this is true, perhaps the changes observed in the
application of categories to people might result from a variety of different
cognitive or social processes or a particular combination of them.
We know that older children are more sophisticated than younger children
in their judgments of people, including themselves. However, I believe we
should be careful not to attribute these changes entirely to cognitive devel-
opment. Presumably a portion of this developmental change may result from
a lengthy period of exposure to a developmentally ordered social environ-
ment (Bronfenbrenner, 1999). For example, Ruble et al. refer to the “deeper
meanings” taken on by gender and race in mid childhood. There must be,
surely, a relationship between the meanings that are conveyed within a child’s
environment and those that the child makes use of. The behaviour of other
children and of adults, the content of mass media directed towards children,
and even the physical accessibility of objects in the world (determined by
strength, height, freedom to roam, etc.) must all impinge on the meanings
available to a child. It seems there is a risk that this point is taken for granted
rather than analysed directly by social identity researchers.
Ruble et al. focus on the ages 7–10 years as “ key period” for studying the
incorporation of social category memberships into the emerging self-concept.
This does seem to fit the evidence, but perhaps a similar argument could be
made for other age periods? For example, for identical twins perhaps the ages
of 1–3 years would be critical in the formation of a “we” unit (cf. Nicole &
Del Miglio, 1995). Equally, adolescence is arguably a period in which peer-
group affiliations become more established and in which the question “what
sort of person am I?” may be central to which social norms are relevant
for the self-concept. Perhaps “key periods” can be identified at many age
boundaries because the social milieu changes externally depending on one’s
age. Imagine yourself as a 2-, 5-, 8-, 12-, 16-, 20-, or 25-year-old today. Each
of these ages is likely to be found inhabiting a distinct social environment
(see Alfieri, Ruble, & Higgins, 1996, for an example of an environment-based
age-related change). It might be argued, perhaps in evolutionary terms, or
perhaps on the basis of cognitive and physical capacity, that these environ-
 Abrams 293
ments are somehow tailored to meet the particular needs of each age group.
However, historical changes in age-related social environments (and indeed in
the self, see Baumeister, 1999) suggest that is likely to be only part of the
story. For example, it is less than 50 years ago that children in Britain might
have left school at 14 to work in the mills, the mines, or on the farm. At the
start of the 21st century the majority remain in the education system until
well past 18 years of age. These rapid changes in societal arrangements might
imply that the mission of capturing critical moments in the development of
social identity is likely to falter because those moments involve combinations
of both cognitive and social factors, and either or both may be absent or
present at various times. The slipperiness of critical developmental stages in
social identity development is also illustrated throughout the book in the
evidence presented by Bennett and Sani, Verkuyten, Barrett et al., and
Rutland.
Ruble et al. and Powlishta note that gender awareness emerges by the age
of 2–3 years, whereas racial and ethnic awareness emerge later. This lag may
point to a difference in the functional value of these categories rather than
just a cognitive advantage for gender. Ruble’s question as to whether, between
3 and 5 years of age, ethnic self-labelling “means anything more than empty
labels”, invites a further set of questions. Presumably, group labels acquire
meaning if they signify a difference or similarity in relation to some other
object or person. We have to ask: Why are such labels used? A likely reason is
that they are a meaningful concept in communication with others (Lyons &
Kashima, 2001; Stangor, Sechrist, & Jost, 2001). With use, experience, and
relevance, that meaning is likely to increase in richness. For example, group
differences in ethnic constancy should encourage us to examine the role
of the social environment. Why would racial constancy lag behind for
African-American and Asian-American children? Ruble et al. propose
that African-American children need to be more flexible about race as they
face the challenge of identifying with both their own racial and ethnic group
and the larger European-American society, or that African-American
children disidentify. An alternative possibility is that precisely the opposite
happens. African-American children may live in a more racially homo-
geneous environment than their European-American counterparts. Therefore
it may be less meaningful to refer to race and ethnicity because it is a homo-
geneous background to their social world. In certain circumstances, being
non-white may not be remarkable (W. J. McGuire & McGuire, 1988; W. J.
McGuire, McGuire, Child, & Fujioka, 1978). In contrast, European-
Americans may experience other races as salient and distinctive, and
therefore it may be more meaningful to note the differences.
How closely might we expect social categorisation and self-categorisation
to reflect one another? Again, the role of the social environment seems crit-
ical. As Ruble et al. note, “there are wide individual and group variations in
how salient various social categories are to the self-concept. . . . Future
research must examine the moderating influence of context”. Of course, the
294 The development of social identity
context can be defined either broadly (culture, historical period) or specific-
ally (as in the comparative context). Distinctions between the implications of
these different aspects of context may be something to which we need to pay
more attention.
Ruble et al. also invoke the notion of centrality, perhaps the chronic sali-
ence, of identity. From a self-categorisation perspective these two are related
simply because the more often an identity is salient the more accessible it will
be and (presumably) the better it will fit a regularly encountered environment.
The finding by Alvarez, Ruble, and Bolger (2001) that the son/daughter
identity was more central than other social identities illustrates this point. On
the other hand, the Alvarez et al. study does not seem to illuminate whether,
in a different context (e.g., the playground), other social identities might dom-
inate (Strough & Marie-Covatto, 2002). Some further theoretical work is
probably required regarding the extent to which centrality may describe a
structural feature in the self-concept or whether it involves contextual
centrality in the process of self-conception (see also David et al., Chapter 5).
Ruble et al. place weight on the role of evaluation in the development of
social identity. The motivational assumptions of social identity theory have
been interpreted by many researchers as a drive to sustain self-esteem. There
are various mechanisms and processes that appear to operate in adults that
result in a relatively automatic bias towards self (cf. Otten & Wentura, 1999),
and these seem to operate outside conscious awareness (Perdue, Dovidio,
Gurtman, & Tyler, 1990). However, the preoccupation in North American
research with the potential low self-esteem of minority group members
should not be taken as a paradigm for analysing the social identity of mem-
bers of all social categories (cf. Emler 2001; Heine, Lehman, Markus, &
Kitayama, 1999). Motivation for self-esteem could have numerous con-
sequences, only one of which is in-group favouritism or bias (Abrams, 1992;
Abrams & Hogg, 1988, 2001). Moreover, it is not at all clear that in-group
favouritism is always motivated by need for self-esteem, even if it results in
changes in self-esteem (Abrams & Hogg, 1988; Stangor & Thompson, 2002).
Many forms of social differentiation do not involve bias or prejudice. For
example, in a school, children and adults know quite clearly that the “staff
room” is for adults. It seems unlikely that observance of this rule affects the
self-esteem of either group. Similarly, girls and boys may be assigned different
washrooms. Use of the designated room is unlikely to have self-esteem impli-
cations (although entering the wrong room may well cause embarrassment).
Instead of the self-esteem principle, Michael Hogg and I (Abrams, 1992;
Abrams & Hogg, 1988, 2001; Hogg, 2001; Hogg & Abrams, 1993) proposed
that a motivational priority is to establish meaning – to make sense of one’s
location in a social environment. This is a prerequisite for engagement in
actions that are socially appropriate and effective and that can be under-
stood by others. It also reduces subjective uncertainty, which may be a
psychologically satisfying state (see also Kruglanski & Webster, 1996).
A further problem arises in the operationalisation of self-esteem.
 Abrams 295
Researchers have a tendency to interpret and measure self-esteem in personal
rather than collective terms. However, in SIT the person–group distinction is
just as relevant to self-esteem as it is to identity (Tajfel, 1974) and behaviour
(cf. Runciman, 1966). It is explicit that groups and individuals may be judged
differently on different dimensions, and that there may even be a dynamic
relationship among these evaluations. Assessments of the in-group bias–
self-esteem link too often used global measures of self-esteem (e.g., feeling
positive about the self overall, or the group as a whole, or possibly all groups
to which one belongs) rather than focusing on a specific social identification
and dimensions of comparison (e.g., feeling good about the specific group’s
academic status, team performance, cultural vitality). For SIT it may be that
the act of evaluating an in-group more highly than an out-group on a specific
dimension is the enhancement of social identity (see Abrams & Hogg, 1988;
Rubin & Hewstone, 1998).
Despite these cautionary notes, there are reasons to study personal self-
esteem in intergroup situations. Indeed, Gaertner, Sedikides, Vevea, and
Iuzzini (2002) propose that the individual self has motivational primacy over
the collective and contextual self. Given the evidence that personality appears
to comprise a considerable element of genetically determined stability (e.g.,
Hur, McGue, & Iacono, 1998; McGue, Bacon, & Lykken, 1993; S. McGuire,
Neiderhiser, Reiss, Hetherington, & Plomin, 1996), it is not surprising that
there is also self-generated stability in self-esteem. Along with this stability
there is also reliable variation between individuals that can be attributed to
different types of environment (e.g., Caspi, Taylor, Moffitt, & Plomin, 2000).
Indeed, self-concept stability may be partly associated with environmental
stability but over time provides a gradually “moving baseline” against which
there are situational fluctuations (Demo, 1992; cf. Ardelt, 2000). Irrespective
of the weight of evidence surrounding the person-situation and environment-
gene debates, what seems clear is that all of these variables contribute to
personality and a host of other socially expressed attributes. There is also a
strong case that a psychological self-structure develops that is relatively
stable, at least in terms of its motivational components (cf. Higgins, 1989). It
is therefore not surprising to find evidence in the adult literature that global
personal self-esteem is related to in-group bias (Aberson, Healy, & Romero,
2000). Similarly, the evidence described by Powlishta shows that children’s
personal self-evaluation rises in line with gender in-group favouritism. More-
over, the fact that the self-esteem–bias relationship is stronger among
girls than boys suggests that variables other than social identification may
underpin this male–female difference.

Sani and Bennett

Sani and Bennett propose that children’s knowledge of group identities
moves through age-related stages, such that they only begin to focus on
beliefs and attitudes related to identity quite late on. It is encouraging that
296 The development of social identity
children aged 12 years or so are sensitive to sociopolitical issues that might
underpin perceptions of shared beliefs. This is also reflected in the inclusion
of these issues as part of the school curriculum for older children, but not
younger children. More intriguing is that older children have developed a
framework or vocabulary for describing the impact of values and beliefs on
behaviour even when they encounter new intergroup contexts. Children as
young as 5 years of age appear to be able to make a clear connection between
psychological attributes and category memberships, but only older children
have more extensive notions of the principles and ideas underlying commit-
ment to an identity. As Sani and Bennett point out, the understanding of
belief-based differences is more central to some category memberships than
others, and there may be important questions of domain-relevant expertise
that need to be understood. Although Sani and Bennett point to the fact that
social identities differ (following Deaux, Reid, Mizrahi, & Cotting, 1999),
perhaps an equally relevant point is that social identification can involve
different types of content depending on a variety of factors. Some of these
are cognitive, but an important set may be found in the social context, the
nature of particular relationships, the relative power and status of different
social groups, and the child’s active engagement as a self-directed agent in
their social world (Abrams, 1992).
Sani and Bennett, and David et al., are enthusiastic about the idea that
social identity (and stereotyping and group preference) is flexible. Sani and
Bennett suggest that since contextual variability in stereotyping emerges at 7
years of age, “once applied to the in-group, stereotype variability is con-
ceptually equivalent to social identity variability”. Although this proposition
is attractive theoretically, perhaps we need to be cautious in predicting the
cognitive and motivational connections between group evaluations, group
stereotypes, and the self-concept.
In their Study 5, Sani and Bennett asked children to make comparative
judgments between age and gender categories (not to rate stereotypes per se),
and found that children were sensitive to relevant dimensions for category
differentiation. This may demonstrate stereotype flexibility, as Sani and
Bennett conclude. However, an additional possibility might be that children
are skilled in highlighting the most relevant features within a context. Their
apparent flexibility in selecting these features may demonstrate that they pos-
sess reasonably well-developed, and quite stable, representation of a variety
of different stereotypes (see Devine, 1989; Monteith, Ashburn-Nardo, Voils,
& Czopp, 2002, for distinctions between stereotype activation and applica-
tion). Moreover, the finding that boys and girls appear to differ in their levels
of differentiation suggests that the processes at work are in part social, and
not entirely cognitive or developmental. The existence of cultural differences
in the scope and variability of self-descriptions across situations also
supports this point (see also Kanagawa, Cross, & Markus, 2001).
Social identity theory proposes that people are motivated to evaluate their
social identity positively. The link between group membership and social
 Abrams 297
identity is made through the evaluation of the in-group. One way to do this is
to create a stereotypical advantage for the in-group, but there are numerous
other ways that do not rely on existing stereotypes (e.g., the use of social
creativity strategies, or simple in-group favouritism).
Turning from stereotypes to identity, evidence on the linkage between
evaluation and identity in adults does not always suggest a very “flexible” set
of phenomena. As mentioned earlier, among adults, Gaertner et al.’s (2002)
meta-analytic evidence shows that in response to certain threats and
opportunities for enhancement the individual self has motivational primacy
over the collective self and the contextual self. Moreover, work by Deaux and
others (see Abrams, 1992; Deaux, 1992) indicates that the self-concept has
considerable stability because people develop a psychological commitment to
a particular view of themselves. Thus, it may be important not to assume that
a process that allows the self to be defined flexibly requires that the self
structure is entirely a function of social context (Abrams, 1996, 1999). For
example, in Sani and Bennett’s Study 7 there are some curious results. Con-
sistent with an SCT prediction, male children’s self-descriptions became more
stereotypical in the intergroup than in the self-only condition. However,
there was not a qualitative alteration in their self-descriptions. In addition,
female children did not reveal so much difference as a function of condition,
suggesting some imperviousness to intergroup (gender) context.
The idea that “it is not until around 7 years of age at the earliest that
children internalise particular social identities” is also worthy of further
research. This conclusion may be apt for situations in which there is no
obvious conflict between groups. However, just as evidence from the adult
literature shows that increased conflict and other factors can make social
identity more salient and impactful, it seems likely that under some circum-
stances children may internalise social identifications at a younger age. For
example, perhaps young Israeli and Palestinian children have (at least) a rudi-
mentary sense of their identity within the intergroup context (Zecharia, 1990).
In other words, it seems likely to me, and on the basis of previous evidence
on gender and ethnicity, that there is an interaction between the strength
and meaningfulness of social categories in the organisation of daily life, and
the extent to which children internalise group memberships early or late.
Another potentially useful direction for research is implied in the final
paragraph of Sani and Bennett’s chapter, i.e., that it is “fruitful to conceive of
the origins of social identities in terms of co-construction (i.e., between nov-
ice and elder), rather than in terms of purely individual cognitive construc-
tion”. I agree with this, and would amplify it. Co-construction may happen
within a peer group, or between elder and novice, or family members, or
within a church, or between school and child. The social-psychological task is
to make sense of a social context or situation in order to engage with it.
This requires a system of social consensus (cf. Berger & Luckman, 1971;
Moscovici, 1976; Stangor et al., 2001). Once children are able to share in a
social consensus they can move beyond egocentrism (cf. Mead, 1934).
298 The development of social identity
Gender

Powlishta
Powlishta provides an excellent review of the evidence for male/female
relations in childhood, and proposes that generic intergroup processes are at
work. Powlishta wonders why gender stereotypes persist in the face of the
apparent absence of objective differences (Maccoby, 1998). Maybe this is not
such a puzzle, though. First, even when distributions are highly overlapping,
persistent differences are likely to become highly apparent across multiple
comparisons (e.g., the height of North American and British adults is appar-
ent to most people even though the distributions are highly overlapping).
More centrally, some of the evidence cited by Powlishta relates to reactions to
transgressions against behavioural norms rather than against stereotypes.
There is little question that many of the behavioural norms applying to males
and females remain distinct (e.g., dress codes, etiquette, child-bearing; see
Glick & Fiske, 2001, Glick et al., 2000).
The distinction between norms and stereotypes is likely to be helpful
because it may be that children focus more closely on how boys and girls
should behave than on complexities of the underlying traits. Indeed, the same
trait (e.g., friendliness) could be manifested quite differently depending on
prevailing social norms. That is, both boys and girls might consider them-
selves to be stereotypically friendly, but at certain ages friendliness might
involve befriending only own-gender children. As Powlishta, Sen, Serbin,
Poulin-Dubois, and Eichstedt (2001) observe, there is not merely a linkage
between attributes and gender categories but an evaluative response to each
association. It seems likely that social norms are highly relevant in defining
how stereotypes are to be applied. This might imply that investigation of the
development of social identity could benefit as much from studying children’s
sensitivity to social norms as from researching their stereotypes (Abrams,
Rutland, Cameron, & Marques, in press).
Powlishta proposes that males may start with a status advantage in the
early years (cf. Glick & Hilt, 2000) and that girls’ negative reactions to boys
may reflect their resentment over lower status. However, other evidence (e.g.,
see Abrams, 1988; Abrams, Sparkes, & Hogg, 1985; David et al., Chapter 5;
Eccles, 1985) shows that girls have the advantage in terms of competency and
performance in school. This suggests an alternative interpretation in terms of
girls’ awareness and evaluations, reflecting that they have higher status within
the classroom (cf. Mullen, Brown, & Smith, 1992). Other questions posed by
Powlishta, such as why boys show more gender segregation, are also import-
ant. One possibility is that girls, as the more powerful group, actually show
greater control over and variability in their behaviour (cf. Guinote, Judd, &
Brauer, 2002). Thus boys may be denied the opportunity to behave in flexible
(and less masculine) ways, whereas girls are not. The intriguing question is
how this power differential reverses itself in adulthood.
 Abrams 299
Powlishta also suggests that, “attention to gender also may become less
important as gender-role expertise is developed”. While it may well be true
that children become less concerned with simple intergroup differentiation
with age, surely children become more sensitised to gender and, in particular,
concerned with the nature of the potentially romantic and sexualised rela-
tionships between males and females (Montgomery & Sorell, 1998), a trend
that can have negative academic consequences for girls (Brendgen, Vitaro,
Doyle, Markiewicz, & Bukowski, 2002; Neeman, Hubbard, & Masten, 1996).
Children are likely to become increasingly interested in media images that
depict men and women as romantic and sexual partners (Livingstone &
Bovill, 2001). This implies that they might differentiate more strongly with
age, but that the biases they show become more subtle, specific, and, in some
senses, complementary (see Abrams, 1989, for a detailed review).
At the end of the chapter Powlishta wonders about the advantages of
using techniques that might inhibit the use of gender-based categories or
redirect children’s attention to other bases of categorisation. While this
seems laudable in some respects, and reminiscent of Bem’s (1981) call for
people to become “gender aschematic”, it could be problematic in other
respects. Gender pervades most social situations, and is particularly relevant
to the social norms within many. It seems likely that children, and people
generally, learn quickly when and whether to apply gender-based judgments
to their own and others’ actions. Most of the action in the application (and
nonapplication) of gender norms seems likely to reflect social norms rather
cognitive constraints (see Abrams & Masser, 1998; cf. Monteith et al.,
2002).

David et al.
David, Grace, and Ryan’s fundamental assumption is that “since social
contexts are potentially infinite in number, so are the selves that will be
determined by them”. I do not share their confidence that this working
assumption can be easily sustained. Aside from previous debates on this
point in relation to adult’s social environments (see Abrams, 1992, 1996,
1999), children’s social environments are by no means infinitely flexible. They
are externally restricted and are limited in number, essentially revolving
around the home/caring and educational contexts. There is a high degree of
adult control over which environment the child moves into, how, and when.
Along with the lack of environmental flexibility, there does not seem to be
evidence that children’s self-concept wavers greatly (Eder, 1990). Children
aged 18 months seem to show constancy in self-recognition (Campbell,
Shirley, Heywood, & Crook, 2000). So, is social identity infinitely malleable
or it is not? If self-categorisations are “a direct consequence of the genuine
economic, political, and social circumstances of our lives” (David et al.,
Chapter 5), surely they will have the same degree of stability, inertia,
and interdependency as those circumstances provide. As a result, infinite
300 The development of social identity
flexibility may be a theoretical but not empirical possibility (cf. James &
Prout, 1990; Jenks, 2002).
It may be useful to draw a distinction between comparative variation and
flexibility in the self-concept. Ryan and David’s (2002) study shows that
whether people define themselves as connected or independent may vary
according to the comparison group. The David (2000) study shows that, when
given a task by an adult in which age is made relevant, children sort on the
basis of age first and gender second. When only gender is made relevant they
sort by gender. David et al. interpret these results as showing that, “very
clearly . . . the self is not a fixed construct”. However, another way to look at
the findings is that people relate their relatively stable self-concept to the
comparative context in a meaningful way (cf. Canter & Ageton, 1984). It is
perfectly reasonable to describe oneself as introverted compared to a clown
but extraverted compared to a monk, as wise compared to one’s children, but
inexperienced compared to one’s parents. This does not mean one has no
stable sense of one’s level of introversion, wisdom, or experience. The task is
even simpler for objective characteristics. A child may know that he or she is
taller than a sibling, shorter than most adults, and is 1 metre tall. I am
suggesting that comparative variation in self-description does not necessarily
imply an equally flexible self-concept.
David et al. suggest that, “none of the dominant theories provide a ration-
ale for the virulence of the primary school separatist ‘war’, and that self-
categorisation theory provides a compelling account of this”. They propose
that, “when gender identity is salient . . . children will tend to perceive all boys
as the same, all girls as the same, and boys and girls as fundamentally differ-
ent from each other. Girls see the difference as favouring them, while boys see
it as favouring them”. It would be useful to develop these ideas more precisely
because the qualifying words, “tend to” may imply either that some children
may not perceive things in this way, or that all children do but not to a full
extent. David et al. claim that males’ and females’ struggle for positive dis-
tinctiveness explains the “hostility” of primary school gender interactions.
Yet it is not clear that boys and girls see positive distinctiveness as an import-
ant goal, nor that they are particularly hostile to one another (see Hartup &
Abecassis, 2002; Hymel, Vaillancourt, McDougall, & Renshaw, 2002; also
Nesdale, Chapter 8). For example, it seems more likely that instances of
direct hostility will occur within than between groups (Abecassis, Hartup,
Haselager, Scholte, & Van Lieshout, 2002). It is important to explain why,
once gender is made “salient”, boys and girls do not apply gender stereotypes
completely and unambiguously. Why are girls more likely to show in-group
favouritism than boys at one age, but the reverse pattern occurs at another
(e.g., Strough & Marie-Covatto, 2002)?
David et al. suggest that, with puberty, “gender loses its appearance of
‘chronic’ salience, it becomes less important to ‘maximise the difference’,
and gender stereotypes become less prescriptive”. This idea seems at odds
with the possibility that puberty may heighten the salience of gender (and
 Abrams 301
sex), and increase the prescriptiveness of gender-related norms (Abrams, 1989;
Abrams, Marques, Bown, & Henson, 2000; Brendgen et al., 2002; Glick &
Hilt, 2000; Montgomery & Sorell, 1998; Strough & Marie-Covatto, 2002).
David et al. characterise adolescent girls as “internalising the social reality
of sexual inequality and becoming disempowered young women”. It seems
likely, however, that this state may apply to some females but not to all, or
even most. As Condor (1986) and Abrams (1989) pointed out, and as is clear
from pervasive attitudinal variation in forms of sexism (Glick & Fiske, 1996),
people do not necessarily accept traditional sex role ideology or sex stereo-
types. For example, some women identify more than others with other
women, or more precisely with particular types of women. Women may
endorse either traditional or progressive sex role ideologies. The same is true
for men (see Glick & Fiske, 1999). Arguably, then, David et al.’s analysis
would benefit from an acknowledgement that people tend to reflect on the
meaning of their gender identity, and play an active role in how it develops.
Finally, somewhat in line with Powlishta’s thinking, David et al. conclude
that gender is not a special case, but merely an instance of intergroup rela-
tions. This strikes me as doubtful, because it is difficult to think of any other
intergroup relationship that involves pervasive, almost ubiquitous, and
strongly prescribed, intimate relationships between members of the different
groups, nor one in which interdependency between the “groups” is an essen-
tial feature of human survival (see also Katz, 1986). There are some very
specific social practices that allow this to happen, such as arranged marriages,
the informal age gap in heterosexual romantic involvement (which starts with
differential onset of puberty and persists into adulthood), and society’s heavy
reliance on women to bear and raise children. I wholly endorse the idea that
gender must be analysed as a form of intergroup relationship, but I think it
should also be acknowledged that, for various reasons, gender is different
from the general case (Abrams, 1989).

Ethnicity, race, nationality, and the family

Verkuyten
The theme of a normative context for intergroup relations is stressed by
Verkuyten. The interesting point, made very clearly by his research, is that
there are ethno-cultural differences in the way children evaluate their own
and other ethnic groups. This takes us away from the idea of a single devel-
opmental pathway, driven strongly by cognitive processes, and towards the
idea of social pathways that depend on cultural and social practices (Spencer,
1995). As Verkuyten points out, “minority group children may value their
ethnic identity because of supportive relationships with other in-group mem-
bers and because they prefer similarities over differences”. That is to say, the
enduring and meaningful relationships that children have with others within
their own groups shapes the way they understand their ethnicity and ethnic
302 The development of social identity
identity. In contrast to David et al., Verkuyten treats social identity as “a
stable self-structure or trait-like aspect that is expressed independently of a
social situation”. Verykuten’s evidence that differences in ethnic identifica-
tion were greater within than between classrooms does not necessarily imply
that contextual effects are unimportant (classroom may not be a relevant
contextual variable). The within-classroom ethnic composition did affect the
salience of ethnic self-description. However, the correlation between self-
categorisation and self-stereotyping was low (r = .38), which reminds us that
much of the variance in social identity might not be accounted for simply by
self-categorisation. There is more going on, and much of it seems attributable
to culture. This invites further exploration of how cultural values and norms
become embodied in the social identity of children.

Barrett et al.
Much of the research evidence cited in this book presses towards adoption of
an SIT/SCT perspective. Barrett, Lyons, and del Valle’s chapter rings loud
warning bells. They describe what must have been a somewhat frustrating
quest of testing how predictions from SCT and SIT apply to national identity.
Across a wide spectrum of national groups they conclude that the evidence is
very inconsistent (Barrett et al., 1997). Children in all countries tended to
show in-group preference in liking, but varied in terms of evaluations based
on positive and negative attributes. There was little evidence that national in-
group preference and evaluations were related to in-group identification
(aside from among Spanish children, who appear to have a special “social
identity” switch!). In the Barrett, Wilson, and Lyons (1999) study there was
no effect of comparative context on attribution of characteristics to national
groups (see also Sani, Bennett, & Joyner, 1999), no evidence of a relationship
between identification and stereotyping, and no evidence that minority group
children would identify more strongly. Most telling, and in line with Abrams
and Hogg (1988), some children who did not identify strongly with their
national group still showed in-group favouritism. Barrett et al.’s evidence is
also notable for the absence of a consistent developmental pattern. As in
Verkuyten’s chapter, this suggests again the importance of social processes
in the way children (of varying ages) respond to issues of national identity.
Taken together, the evidence seems consistent with the idea that there may be
strong norms at work that may facilitate or inhibit the expression of national-
istic judgments over and above effects of self-categorisation and social iden-
tity, and possibly age. Barrett et al. note that Turner (1999) has recently
accepted that core social identity processes may interact with numerous other
factors, meaning that it might be inappropriate to look for correlations
between pairs of variables. Perhaps the myriad of different developmental
and national patterns in Barrett et al.’s data reflect the moderating effects of
several of these factors. In this case, it could be argued that the way forward is
to develop a more complex, multifaceted, and multivariate approach to
 Abrams 303
research national identity. However, this risks creating unmanageably large
and complex research designs, particularly for use with children. In the light
of all their evidence it is tempting to accept Barrett et al.’s conclusion that
SIT/SCT are potentially impervious to empirical evaluation because all
findings can be explained post hoc. The protective layers around the key ideas
do sometimes seem to render SCT and SIT invulnerable to refutation (see
also Abrams, 1992). However, it would be premature to give up the expedition
just because some of the trails are impassable or circular! Perhaps we need to
concentrate on specifying when SIT will and will not be likely to provide a
useful explanation of children’s intergroup attitudes and behaviour.

Nesdale
Nesdale highlights a distinction made by some other contributors between
in-group preference and out-group rejection. He offers social identity
development theory as a way of accommodating this distinction. Nesdale’s
theory offers a clear developmental framework within which social identity
processes may emerge, at least for ethnicity. He outlines several tasks for
such a theory, including accounting for the discrepancy between children’s
attitudes and those of their family members, recognising the developmental
changes in perceptual, cognitive, and linguistic abilities, the operation of
social motivations, and the fact that some children seem not to develop
ethnic prejudices. Along with Rutland, Nesdale notes that SIT and SCT have
nothing particular to say about cognitive abilities.
The first important distinction is that children respond to ethnic self-
categorisation by focusing only on the in-group. The emergence of prejudice,
or active dislike for an out-group, occurs later and depends on children
adopting as their own the prejudices of members of an in-group with
which they identify. Prejudice is thus more likely if the group as a whole is
prejudiced or is in conflict with or threatened by the out-group.
A particularly valuable contribution of Nesdale’s analysis is that it
specifies conditions under which we would not expect children to express
prejudice. As he notes, “young children up to 9 years of age are simply not
repositories of ethnic dislike and prejudice” (see also Hartup & Abecassis,
2002). Only when the context and motivation are pressing towards prejudice
is it likely to appear. Similar points can be made in relation to a range of
social categorisations, including age, gender, and nationality.
Nesdale shows how a developmental approach can be combined with SCT
and SIT. This approach adopts a stage-based analysis, which implies cogni-
tive and/or social limits on what children might do in relation to social
categories at different ages. It would be nice to see this framework extended.
The competence/performance distinction is certainly relevant here, as are a
host of interesting questions. What are the developmental steps in children’s
ability to judge the relationships between social attributes and social cate-
gories that do and do not include themselves? Do the same categorisation
304 The development of social identity
processes (e.g., metacontrast) operate as in adulthood, and can we simply
draw the regression line back from adulthood to childhood? At what age does
the metacontrast process emerge? What precedes it? When does depersonal-
isation start to happen? What precedes it? What are the initial motivations
that accompany self-categorisation? Is there an automatic desire for positive
distinctiveness or is this a culturally determined motive (Heine et al., 1999)?
During childhood, are the changes in social identification and intergroup
behaviour reflective of qualitative changes (in process) or merely a gradual
quantitative change that culminates in adult-like behaviour? How does the
development of other social competencies combine with the development
of categorisation processes? Can the development of social identity tell us
something new about the processes that might operate in adults? Why do
some children who apparently share the same social environment display
different intergroup attitudes and behaviour? Are these differences attri-
butable to cognitive differences or are there powerful social factors beyond
the immediate context that must be better understood? By pointing to the
fact that children’s ethnic identity does not conform to the stereotypical
intergroup template, Nesdale reminds us that the incorporation of SIT and
SCT in a developmental framework requires additional theorising about both
development and social identity processes.

Rutland
Given my interest in self-regulation in children (Abrams, 1984, 1985) and
normative processes generally (Abrams & Hogg, 1990), I welcome Rutland’s
exploration of the way children self-regulate their intergroup attitudes.
Indeed, the issue of self-presentation is surely relevant for children in most
experimental studies, within which they are implicitly or explicitly account-
able to an adult for their behaviour. Introducing the use of implicit attitude
measurement to the area of children’s social identity, Rutland reports
evidence that explicit biases may decrease with age but implicit ones may not.
Children’s moderation of public expressions of prejudice is an important
phenomenon. It indicates that they understand prevailing norms regarding
the expression of intergroup attitudes. It also implies that they may be more
motivated to accommodate to those norms than to serve their social identity.
The more general point is that the social environment within which children
learn ethnic, gender, and national attitudes may contain powerfully internal-
ised norms that affect their expression of prejudice. In turn, against a politic-
ally liberal background this may restrict the levels of prejudice that actually
exist (see Nesdale’s theory, and also Stangor et al., 2001).
Just as norms may vary, so will the impact of norms on prejudice and
behaviour. Presumably the children of Hutus and Tutsis in Rwanda
developed profound intergroup hatred from a very early age, reinforced by
extremely strong social norms. The point is that we need to build an under-
standing of how children learn and embrace such norms into our theorising
 Abrams 305
about the development of social identity and patterns of prejudice. Given the
current interest in social self-regulation of prejudice in adults (Abrams, 1990,
1994; Abrams & Masser, 1998; Monteith et al., 2002; Plant & Devine, 1998),
it seems that this area is well-worth pursuing in the analysis of children’s
prejudice.

Banker et al.
Banker, Gaertner, Dovidio, Houlette, Johnson, and Riek expand the context
of childhood into the realm of families. This chapter is less concerned with
the cognitive-developmental processes within individuals than with the way
families may be characterised by different intergroup relations. Banker et al.
find that the perception of the stepfamily as part of a common group does
much to reduce conflict and facilitate positive relationships.
The idea of viewing stepfamilies in group terms raises interesting possi-
bilities for promoting harmony within families in general. In particular, there
may be several ways that families could be partitioned into different groups
(e.g., generational groups, geographical groups). How would different models
of intergroup contact propose that family relationships could be sustained
most positively? Pettigrew’s (1998) reformulated contact model proposes that
positive intergroup attitudes are most likely to develop if initial contact is
depersonalised, intergroup differences are subsequently attended to, and
finally a superordinate identity is recognised. Perhaps this sort of sequence is
a natural aspect of widening one’s family (e.g., the addition of in-laws). Other
approaches (e.g., Hewstone & Brown, 1986; Hornsey & Hogg, 2000) suggest
that a dual identity strategy is likely to succeed, in which people are able to
sustain positive intergroup distinctiveness at the same time as sharing a
common identity.
The application of the common in-group identity model at the level of
families also raises interesting questions about the way children might
develop or inhibit different aspects of social identity. For example, how can a
child who sees his or her parents and self as a unit also differentiate strongly
between men and women or boys and girls? Is the family unit the primary
group for most children? If so, why has the study of children’s intergroup
perceptions ignored this unit in preference to other group memberships?
Given that family is often a central group for collectivist cultures (cf.
Trafimow, Triandis, & Goto, 1991), perhaps there is more to be learnt here.

CONCLUSIONS

My aim in this chapter has been to highlight some of the ideas and
findings in this book that will provide a spur for debate and further
research. Although I have focused on issues that are debatable, my overall
sense is that the book portrays a relatively coherent and consensual view
306 The development of social identity
of what research can tell us about the development of social identity. Clearly
the social identity approach offers plenty of interesting research questions for
developmentalists, and vice versa.
At the outset of this chapter I asked whether the cognitive-motivational
elements of self-categorisation and social identity should be the only, or the
most important, aspect of children’s group membership that should attract
our interest. Are we asking the right questions about children’s relationship
to social groups and categories? I think future research would benefit by
devoting further attention to the social (e.g., cultural, intragroup, and inter-
personal) processes by which intergroup relationships are defined, sustained,
and given continued meaning for children.
Second, the different chapters in this book reveal a certain tension between
social-contextual and cognitive-developmental accounts of social identity in
children. I asked what other elements in the developmental equation may
have been neglected. As Bennett and Sani note in their introduction, there
are good reasons to think that social identity involves developmental pro-
cesses across the lifespan as well as between situations, contexts, and inter-
group relationships. Perhaps equally important is to be aware that social
identity is not all in the head. It is something that must acquire shared
meaning, and that necessarily involves social interaction between people and
in relatively stable social environments. Rather little of the work described in
this book examines directly how those interactions may be involved in the
way social identity is constructed, and this could be a very productive avenue
for future research.
Third, there is controversy in the adult literature surrounding various
assumptions made by social identity theory and self-categorisation theory.
How far does developmental research reflect these controversies? There has
been some progress, and a host of interesting questions invite further explor-
ation. For example, these include the way children regulate their behaviour as
group members, the developmental impact of different motivational elements
such as uncertainty reduction, need for cognition, terror management,
impression management, optimal distinctiveness, and self-verification, and
the link between communication and social identity.
To conclude, the important point to emerge from all of the fascinating
work and ideas in this book is that social identity phenomena invite an analy-
sis that goes beyond a cognitive-developmental approach and beyond a
social-cognitive approach. We do need to specify the distinctive effects of
cognitive development and of self-categorisation and social identification.
But to predict how these will be manifested as children grow older almost
certainly requires a more extensive analysis of the way culture, cultural rules
and traditions, and specific group norms are maintained within groups, as
well as an account of the historical context of intergroup relations. The chap-
ters in this book reflect very substantial progress in clearing the ground theor-
etically and methodologically. The book as a whole provides a stimulus for
many interesting and important questions for future research.
 Abrams 307
REFERENCES

Abecassis, M., Hartup, W. W., Haselager, G. J. T., Scholte, R. H. J., & Van Lieshout,
C. F. M. (2002). Mutual antipathies and their significance in middle childhood
and adolescence. Child Development, 73, 1543–1556.
Aberson, C. L., Healy, M., & Romero, V. (2000). In-group bias and self-esteem: A
meta-analysis. Personality and Social Psychology Review, 4, 157–173.
Abrams, D. (1984). Social identity, self-awareness and intergroup behaviour. PhD
thesis, University of Kent, Canterbury, UK.
Abrams, D. (1985). Focus of attention in minimal intergroup discrimination. British
Journal of Social Psychology, 24, 65–74.
Abrams, D. (1988). Sex-role ideology, siblings and educational attainment: Social and
psychological factors. ESRC 16–19 Initiative Working Paper. Swindon: Economic
and Social Research Council.
Abrams, D. (1989). Differential association: Social developments in gender identity
and intergroup relations during adolescence. In S. Skevington & D. Baker (Eds.),
The social identity of women (pp. 59–83). London: SAGE Publications.
Abrams, D. (1990). The self-regulation of group behavour: An integration of
self-awareness and social identity theory. In D. Abrams & M. Hogg (Eds.), Social
Identity theory: Constructive and critical advances (pp. 89–112). Hemel Hempstead:
Harvester Wheatsheaf.
Abrams, D. (1992). Processes of social identification. In G. Breakwell (Ed.), The
social psychology of the self-concept (pp 57-100). London: Academic Press/Surrey
University Press.
Abrams, D. (1996). Social identity, self as structure and self as process. In W. P.
Robinson (Ed.), Social groups and identities: Developing the legacy of Henri Tajfel
(pp. 143–168). London: Butterworth Heinemann.
Abrams, D. (1999). Social identity, social cognition and the self: The flexibility and
stability of self-categorization. In D. Abrams & M. A. Hogg (Eds.), Social identity
and social cognition (pp. 197–229). Oxford: Blackwell.
Abrams, D., & Hogg, M. A. (1988). Comments on the motivational status of self-
esteem in social identity and intergroup discrimination. European Journal of Social
Psychology, 18, 317–334.
Abrams, D., & Hogg, M. A. (1990). Social identification, self-categorisation and
social influence. In M. Hewstone & W. Stroebe (Eds.), Review of European social
psychology, Vol. 1 (pp. 195–228). Chichester, UK: John Wiley & Sons.
Abrams, D., & Hogg, M. A. (2001). Collective identity: Group membership and
self-conception. In M. A. Hogg & S. Tindale (Eds.), Blackwell handbook of social
psychology, Vol. 3: Group processes (pp. 425–461). Oxford: Blackwell.
Abrams, D., Marques, J. M., Bown, N. J., & Henson, M. (2000). Pro-norm and anti-
norm deviance within in-groups and out-groups. Journal of Personality and Social
Psychology, 78, 906–912.
Abrams, D., & Masser, B. (1998). Context and the social self-regulation of
stereotyping: Perception, judgment and behavior. In R. S. Wyer (Ed.), Advances in
social cognition, Vol. 11 (pp. 53–67). Hillsdale, NJ: Lawrence Erlbaum Associates,
Inc.
Abrams, D., Rutland, A., Cameron, L., & Marques, J. M. (in press). The development
of subjective group dynamics: When in-group bias gets specific. British Journal of
Developmental Psychology.
308 The development of social identity
Abrams, D., Sparkes, K., & Hogg, M. A. (1985). Gender salience and social identity:
The impact of sex of siblings on educational and occupational aspirations. British
Journal of Educational Psychology, 55, 224–232.
Alfieri, T., Ruble, D. N., & Higgins, E. T. (1996). Gender stereotypes during
adolescence: Developmental changes and the transition to junior high school.
Developmental Psychology, 32, 1129–1137.
Alvarez, J. M., Ruble, D. N., & Bolger, N. (2001). The role of evaluation in the
development of person perception. Child Development, 72, 1409–1425.
Ardelt, M. (2000). Still stable after all these years? Personality stability theory
revisited. Social Psychology Quarterly, 63, 392–405.
Barrett, M., Lyons, E., Bennett, M., Vila, I., Gimenez, A., Arcuii, L., & de Rosa,
A. S. (1997). Children’s beliefs and feelings about their own and other national
groups in Europe. Final Report to the Commission of the European Communi-
ties, Directorate-General Xii for Science, Research and Development, Human
Capital and Mobility (HCM) Programme, Research Network No. CHRX-CT94–
0687.
Barrett, M., Wilson, H., & Lyons, E. (1999, April). Self-categorization theory and the
development of national identity in English children. Poster presented at Biennial
Meeting of SRCD, Alberquerque, USA.
Baumeister, R. F. (Ed.) (1999). The self in social psychology. Philadelphia, PA:
Psychology Press.
Bem, S. L. (1981). Gender schema theory: A cognitive account of sex-typing.
Psychological Review, 88, 354–364.
Berger, P. L., & Luckman, T. (1971). The social construction of reality. Harmondsworth,
UK: Penguin.
Brendgen, M., Vitaro, F., Doyle, A. B., Markiewicz, D., & Bukowski, W. M. (2002).
Same-sex peer relations and romantic relationships during early adolescence:
Interactive links to emotional, behavioral and academic adjustment. Merill Palmer
Quarterly, 48, 77–103.
Bronfenbrenner, U. (1999). Environments in developmental perspective: Theoretical
and operational models. In S. L. Friedman & T. D. Wachs (Eds.), Measuring
environment across the lifespan: Emerging methods and concepts (pp. 3–28).
Washington, DC: American Psychological Association.
Campbell, A., Shirley, L., Heywood, C., & Crook, C. (2000). Infants’ visual preference
for sex-congruent babies, children, toys and activities: A longitudinal study. British
Journal of Developmental Psychology, 18, 479–498.
Canter, R. J., & Ageton, S. S. (1984). The epidemiology of adolescent sex-role
attitudes. Sex Roles, 11, 657–676.
Caspi, A., Taylor, A., Moffitt, T. E., & Plomin, R. (2000). Neighborhood deprivation
affects children’s mental health: Environmental risks identified in a genetic design.
Psychological Science, 11, 338–342.
Condor, S. G. (1986). Sex role beliefs and “traditional” women: Feminist and inter-
group perspectives. In S. Wilkinson (Ed.), Feminist social psychology (pp. 97–118).
Milton Keynes, UK: Open University Press.
Deaux, K. (1992). Personalizing identity and socializing self. In G. Breakwell (Ed.),
Social psychology of identity and the self-concept (pp. 9–34). London: Academic
Press.
Deaux, K., Reid, A., Mizrahi, K., & Cotting, D. (1999). Connecting the person to the
social: The functions of social identification. In T. R. Tyler, R. M. Kramer, & O. P.
 Abrams 309
John (Eds.), The psychology of the social self: Applied social research (pp. 91–113).
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Demo, D. H. (1992). The self-concept over time: Research issues and directions.
Annual Review of Sociology, 18, 303–326.
Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled
components. Journal of Personality and Social Psychology, 56, 5–18.
Eccles, J. (1985). Sex differences in achievement patterns. In R. A. Dienstbier &
T. B. Sonderegger (Eds.), Nebraska Symposium on Motivation, 1984 (pp. 97–132).
Lincoln, NE: University of Nebraska Press.
Eder, R. A. (1990). Uncovering young children’s psychological selves: Individual and
developmental differences. Child Development, 61, 849–863.
Emler, N. P. (2001). Self-esteem: The costs and causes of low self-worth. York, UK:
Joseph Rowntree Foundation/York Publishing Services.
Gaertner, L., Sedikides, C., Vevea, J. L., & Iuzzini, J. (2002). The “I”, the “we” and
the “when”. A meta-analysis of motivational primacy in self-definition. Journal of
Personality and Social Psychology, 83, 574–591.
Glick, P., & Fiske, S. T. (1996). The ambivalent sexism inventory: Differentiating
hostile and benevolent sexism. Journal of Personality and Social Psychology, 70,
491–512.
Glick, P., & Fiske, S. T. (1999). The ambivalence toward men inventory: Differentiat-
ing hostile and benevolent beliefs about men. Psychology of Women Quarterly, 23,
519–536.
Glick, P., & Fiske, S. T. (2001). Ambivalent sexism. In M. P. Zanna (Ed.), Advances in
experimental social psychology, Vol. 33 (pp. 115–188). San Diego, CA: Academic
Press.
Glick, P., Fiske, S. T., Mladinic, A., Saiz, J. L., Abrams, D., Masser, B., Adetoun, B.,
Osagie, J. E., Akande, A., Alao, A., Anneteje, B., Willmsen, T. M., Chipeta, K.,
Dardenne, B., Dijsterhuis, A., Wigboldus, B., Eckes, T., Six-Materna, I., Exposito,
F., Moyal, M., Foddy, M., Kim, H. J., Lameiras, M., Sotelo, M. J., Micchi-Faina,
A., Romani, M., Sakalli, N., Udegbe, B., Yamamoto, M., Ui, M., Ferreira, M. C.,
& Lopez, W. L. (2000). Beyond prejudice as simple antipathy: Hostile and
benevolent sexism across cultures. Journal of Personality and Social Psychology, 79,
763–775.
Glick, P., & Hilt, L. (2000). Combative children to ambivalent adults: The develop-
ment of gender prejudice. In T. Eckes & H. Traunter (Eds.), The developmental
social psychology of gender (pp. 243–272). Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Guinote, A., Judd, C. M., & Brauer, M. (2002). Effects of power on perceived and
objective group variability: Evidence that more powerful groups are more variable.
Journal of Personality and Social Psychology, 82, 708–721.
Hartup, W. W., & Abecassis, M. (2002). Friends and enemies. In P. K. Smith & C. H.
Hart (Eds.), Blackwell handbook of childhood social development (pp. 285–306).
Oxford: Blackwell.
Heine, S. H., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a
universal need for positive self-regard? Psychological Review, 106, 766–794.
Hewstone, M., & Brown, R. J. (1986). Contact is not enough: An intergroup perspec-
tive on the “contact hypothesis”. In M. Hewstone & R. J. Brown (Eds.), Contact and
conflict in intergroup encounters: Social psychology and society (pp. 1–44). Oxford:
Blackwell.
310 The development of social identity
Higgins, E. T. (1989). Self-discrepancy theory: What patterns of self-beliefs cause
people to suffer? In L. Berkowitz (Ed.), Advances in experimental social psychology,
Vol. 22 (pp. 93–136). San Diego, CA: Academic Press.
Hogg, M. A. (2001). Self-categorization and subjective uncertainty resolution:
Cognitive and motivational facets of social identity and group membership. In
J. P. Forgas, K. D. Williams, & L. Wheeler (Eds.), The social mind: Cognitive and
motivational aspects of interpersonal behavior (pp. 323–349). New York: Cambridge
University Press.
Hogg, M. A., & Abrams, D. (1993). An uncertainty reduction model of group motiv-
ation. In M. A. Hogg & D. Abrams (Eds.), Group motivation: Social psychological
perspectives (pp. 173–190) London: Harvester Wheatsheaf.
Hornsey, M. J., & Hogg, M. A. (2000). Subgroup relations: A comparison of mutual
intergroup differentiation and common ingroup identity models of prejudice
reduction. Personality and Social Psychology Bulletin, 26, 242–256.
Hur, Y.-M., McGue, M., & Iacono, W. G. (1998). The structure of self-concept in
female preadolescent twins: A behavioral genetic approach. Journal of Personality
and Social Psychology, 74, 1069–1077.
Hymel, S., Vaillancourt, T., McDougall, P., & Renshaw, P. D. (2002). Peer
acceptance and rejection in childhood. In P. K. Smith & C. H. Hart (Eds.), Black-
well handbook of childhood social development (pp. 265–284). Oxford: Blackwell
Publishing.
James, A., & Prout, A. (1990). Constructing and reconstructing childhood. Basingstoke,
UK: Falmer.
Jenks, C. (2002). A sociological approach to childhood development. In P. K. Smith &
C. H. Hart (Eds.), Blackwell handbook of childhood social development (pp. 79–96).
Oxford: Blackwell.
Kanagawa, C., Cross, S. E., & Markus, H. R. (2001). “Who am I?” The cultural
psychology of the conceptual self. Personality and Social Psychology Bulletin, 27,
90–103.
Katz, P. A. (1986). Gender identity: Development and consequences. In
R. D. Ashmore & F. K. Del Boca (Eds.), The social psychology of female–male
relations: A critical analysis of central concepts (pp. 21–68). New York: Academic
Press.
Kruglanski, A. W., & Webster, D. M. (1996). Motivated closing of the mind: “Seizing”
and “freezing”. Psychological Review, 103, 263–283.
Livingstone, S., & Bovill, M. (Eds.) (2001). Children and their changing media environ-
ment: A European comparative study. Mahwah, NJ: Lawrence Erlbaum Associates,
Inc.
Lyons, A., & Kashima, Y. (2001). The reproduction of culture: Communication
processes tend to maintain cultural stereotypes. Social Cognition, 19, 372–394.
Maccoby, E. E. (1998). The two sexes: Growing up apart, coming together. Cambridge,
MA: Harvard University Press.
McGue, M., Bacon, S., & Lykken, D. T. (1993). Personality stability and change
in early adulthood: A behavioral genetic analysis. Developmental Psycholoy, 29,
96–109.
McGuire, S., Neiderhiser, J. M., Reiss, D., Hetherington, E. M., & Plomin, R. (1996).
“Genetic and environmental influences on perceptions of self-worth and com-
petence in adolescence: A study of twins, full siblings and step-siblings:” Erratum.
Child Development, 67, 3417.
 Abrams 311
McGuire, W. J., & McGuire, C. V. (1988). Content and process in the experience of the
self. In L. Berkowitz (Ed.), Advances in experimental social psychology, Vol. 21.
Social psychological studies of the self: Perspectives and programs (pp. 97–144). San
Diego, CA: Academic Press.
McGuire, W. J., McGuire, C. V., Child, P., & Fujioka, T. (1978). Salience of ethnicity
in the spontaneous self-concept as a function of one’s ethnic distinctiveness in the
social environment. Journal of Personality and Social Psychology, 36, 511–520.
Mead, G. H. (1934). Mind, self and society. Chicago, IL: University of Chicago
Press.
Monteith, M. J., Ashburn-Nardo, L., Voils, C. I., & Czopp, A. M. (2002). Putting the
brakes on prejudice: On the development and operation of cues for control. Journal
of Personality and Social Psychology, 83, 1029–1050.
Montgomery, M. J., & Sorell, G. T. (1998). Love and dating experience in early and
middle adolescence: Grade and gender comparisons. Journal of Adolescence, 21,
677–689.
Moscovici, S. (1976). Social influence and social change. London: Academic
Press.
Mullen, B., Brown, R.J., & Smith, C. (1992). Ingroup bias as function of salience,
relevance and status: An integration. European Journal of Social Psychology, 22,
103–122.
Neeman, J., Hubbard, J., & Masten, A. S. (1996). The changing importance of roman-
tic relationship involvement to competence from late childhood to late adolescence.
Development and Psychopathology, 7, 727–750.
Nicole, S., & Del Miglio, C. (1995). Self-awareness in monozygotic twins: A relational
study. Acta Geneticae Medicae et Gemellologiae: Twin Research, 44, 107–115.
Otten, S., & Wentura, D. (1999). About the impact of automaticity in the Minimal
Group Paradigm: Evidence from affective priming tasks. European Journal of
Social Psychology, 29, 1049–1071
Perdue, C. W., Dovidio, J. F., Gurtman, M. B., & Tyler, R. B. (1990). Us and them:
Social categorization and the process of intergroup bias. Journal of Personality and
Social Psychology, 59, 475–486.
Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49,
65–85.
Plant, E. A., & Devine, P. G. (1998). Internal and external motivation to respond
without prejudice. Journal of Personality and Social Psychology, 75, 811–832.
Powlishta, K. K., Sen, M. G., Serbin, L. A., Poulin-Dubois, D., & Eichstedt, J. A.
(2001). From infancy through middle childhood: The role of cognitive and social
factors in becoming gendered. In R. K. Unger (Ed.), Handbook of the psychology of
women and gender (pp. 116–132). New York: John Wiley & Sons.
Rubin, M., & Hewstone, M. (1998). Social identity theory’s self-esteem hypothesis. A
review and some suggestions for clarification. Personality and Social Psychology
Review, 2, 40–62.
Runciman, W. G. (1966). Relative deprivation and social justice. London: Routledge &
Kegan Paul.
Ryan, M., & David, B. (2002). A gendered self or a gendered context: A self-
categorization approach to independence and interdependence. Manuscript submitted
for publication.
Sani, F., Bennett, M., & Joyner, L. (1999). Developmental aspects of contextual
variability in the stereotype of the ingroup: The case of Scottish children. Paper
312 The development of social identity
presented at the Annual Conference of the Social Psychology Section of the British
Psychological Society, Lancaster, UK.
Spencer, M. B. (1995). Cultural cognition and social cognition as identity correlates
of black children’s personal-social development. In M. B. Spencer, G. K. Brookins,
& W. R. Allen (Eds.), Beginnings: The social and affective development of black
children: Child psychology (pp. 215–230). Hillsdale, NJ: Lawrence Erlbaum
Associates, Inc.
Stangor, C., Sechrist, G. B., & Jost, J. T. (2001). Changing racial beliefs by providing
consensus information. Personality and Social Psychology Bulletin, 27, 486–496.
Stangor, C., & Thompson, E. P. (2002). Needs for cognitive economy and self-
enhancement as unique predictors of intergroup attitudes. European Journal of
Social Psychology, 32, 563–575.
Strough, J., & Marie-Covatto, A. (2002). Context and age differences in same- and
other-gender peer preferences. Social Development, 11, 346–361.
Tajfel, H. (1974). Social identity and intergroup behaviour. Social Science Informa-
tion, 13, 65–93.
Trafimow, D., Triandis, H. C., & Goto, S. (1991). Some tests of the distinction
between the private self and the collective self. Journal of Personality and Social
Psychology, 60, 649–655.
Turner, J. C. (1999). Some current issues in research on social identity and self-
categorization theories. In N. Ellemers, R. Spears, & B. Boosje (Eds.), Social
identity (pp. 6–34). Oxford: Blackwell.
Zecharia, D. S. (1990). Development of an ethnic self-definition: The ethnic
self-concept “Jew” among Israeli children. International Journal of Behavioral
Development, 13, 317–332.
12 Towards a developmental social
psychology of the social self
Kevin Durkin

Children begin life as members of a social group, or they die. For the human
infant, it is essential to be linked to another person or persons in order to
obtain nutrition, accommodation, basic physical care, and a range of sensory
stimulation matched to innate potentialities. Group membership is a raw
necessity, and nature and cultures have evolved together to ensure that it is
underwritten, universally, by the enduring collectives that we call families.
Children continue their progress into society by joining and identifying
with more groups. At least one of these is of universal significance, highly
salient in all known societies: gender. Others are almost as common and
sometimes imbued with still greater phenomenological consequence: class/
caste, ethnicity, nationality, territory, religion, language group, school. Yet
others are more parochial, though not necessarily less vital to the people
involved in them: friendship groups, sports teams, local child/youth organisa-
tions, gangs, fashions in appearance/music.
With starting points like these, it is not surprising that social affiliations are
important to human beings. It is not surprising that group identities have
motivational, affective, and informational significance. What is surprising is
that developmental psychologists have tended to neglect group identity and
social psychologists have tended to ignore the origins of groups. The many
provocative theoretical and empirical contributions of this volume provide an
overdue correction and demonstrate that exciting research prospects come
into view if we dare to cast off our subdisciplinary blinkers.
I have been asked to comment from a developmental perspective and will
concentrate on issues that relate to how developmental changes come about –
though it will be clear that this entails acknowledging the many ways in which
developmental psychologists can profit from the perspectives of our social
colleagues. I will not reiterate the contents: The editorial introduction pro-
vides an excellent summary and the chapters themselves are impressively
clear. Instead, I will attempt to highlight a number of themes that seem to me
to be interwoven through the chapters, sometimes very explicitly, sometimes
implicitly or tangentially, and perhaps occasionally notable largely because of
their absence. The reader who has followed the text this far will need no
further persuasion that the authors have provided us with many insights and
314 Developmental psychology of the social self
valuable new findings. Hence, I will focus primarily upon issues for future
research on the premise that, for all the advances that it records, the still
greater outlook for this volume is that it will prove to be a catalyst.
Readers will doubtless detect many themes and questions (see also
Abrams, Chapter 11), and the comments below are offered in a catalytic
spirit, that they may contribute to the plentiful reactions the authors
have initiated. For this reader, they lead to the following: (1) the family
as the first group, (2) developmental changes and continuities, (3) language
and the development of social selves, (4) competition between developmental
and social processes, (5) age as a social identity, (6) promoting better social
selves, and (7) social selves through the lifespan. These will be discussed
in turn.

THE FAMILY AS THE FIRST GROUP

To the developmentalist, it seems obvious that the family is important. As

indicated above, the family is crucial to survival and is the gateway to virtu-
ally everything else that children encounter in their early social activities. The
family is the first group (the first “we”) a person joins and it is usually the
longest lasting. To the developmentalist, group processes begin here and
social psychologists appear to be overlooking something fundamental by
disregarding the family or by treating it as just one group among many. On
the other hand, developmentalists tend to pay little attention to the fact
that families stand in some kind of relationship to each other – that is,
that there are in-groups (“my family”) and out-groups (“your family”, “that
family”) – as well as to other levels of community and society. It takes a
social-psychological perspective to begin to see why this could be important.
Several chapters agree that in-group bias associated with social categories
is operating in normally developing children by around 5 or 6 years of age
(Barrett et al., Nesdale, Ruble et al., Rutland, Sani & Bennett). The evidence
is extensive, but perhaps not comprehensive. We have not yet identified the
starting point or the early developmental course of in-group bias. If it is
reasonable to suppose that precursors can be traced to the emotional
embrace of the first group then a more comprehensive account will entail
asking how the powerful processes of attachment and the sustained obliga-
tory location within a social structure (the secure base: Bowlby, 1988), relate
to the feelings and cognitions that children (and older individuals) develop
about other groups.
Family relations appear to be central, in Ruble et al.’s sense of the term, to
many children’s representations of their social selves. In Ruble et al.’s studies
of self-definition, family relations rank alongside gender and ethnicity in a
elicitation task, and ahead of them in a forced-choice task. The family is the
initial context within which children develop the two social identities that
have been most prominent in this volume, namely gender and ethnicity. It
 Durkin 315
becomes clear in Verkuyten’s chapter that ethnic minority, immigrant families
transmit traditions and values that bear on their children’s subsequent inter-
group experiences and attitudes. Furthermore, the family may provide a pro-
tective buffer or alternative source of self-esteem in the face of awareness of
negative social categorisation of one’s ethnic status (Ruble et al., Verkuyten).
It is striking that the editors introduce Banker et al.’s chapter on stepfamily
conflict by noting, quite accurately, that it is a “relatively unusual application
of the social identity approach”. It is unusual, but perhaps it tells us as much
about what is usual in that it seems on first sight quite radical to traverse the
territory between the minimal groups’ laboratory and the realities of inter-
group mergers in the modern family. Yet, as Banker et al. demonstrate, the
conceptual apparatus fits very well and helps explain processes that can go
awry as well as guide interventions designed to promote family cohesion and
well-being.
Verkuyten quotes the perceptive remark of Horowitz that “ethnicity is
family writ large”. We might go on to venture that there is a sense in which
the family is writ large in all of our social selves. This is more than a truism
and less than an explanation: Some aspects of family processes may be
reflected in our orientations to other social categories (emotional affiliation,
in-group bias, deriving self-esteem from membership), and some may be more
indirectly related or even contradictory. Children’s social selves are likely to
become more complex as they are exposed to a new social environment out-
side the family but it is plausible, as implied by Ruble et al., that the family
remains a reference/comparison point for a long time.
Both SIT and SCT maintain that categorisation and internalisation of a
social identity precede the motivational consequences. Turner, Hogg, Oakes,
Reicher, & Wetherell et al. (1987) see group formation as “an adaptive social
psychological process that makes social cohesion, cooperation and influence
possible” (p. 40; and see Bennett & Sani, Chapter 3). Yet social cohesion,
cooperation, and influence are not merely possible but indispensable at the
beginnings of life: We start out with motivations (to survive, to attach, to
explore, to understand) and, for better or worse, our first group is waiting
for us.

DEVELOPMENTAL CHANGES AND CONTINUITIES

Among the fundamental yet most difficult questions facing the broad enquiry
represented here are identifying what the developmental changes are and
specifying when and how they occur. As Bennett and Sani note at the outset,
developmentalists interested in the self have tended to focus on age-related
changes over time, as developmentalists are wont to do. But as we proceed
through the careful and multifaceted investigations in this book, we find
repeatedly that a clear developmental framework is elusive. If at any point we
seem to be getting closer, then context rears its polysemous head.
316 Developmental psychology of the social self
In the course of collaborative research, it is often a disappointment to
nondevelopmental colleagues to learn that developmentalists no longer feel
confident in offering “hard” stage models, delineating exactly what children
can and cannot do at particular age points. This book demonstrates why we
cannot. Certainly, broad cognitive changes can be sketched out. Ruble et al.
highlight the shift from awareness of physical characteristics to the incorpor-
ation of psychological attributes; Sani and Bennett show a development from
awareness of actions and dispositions to understanding of identity-related
beliefs; Nesdale charts development from an undifferentiated phase (with
respect to ethnic categorisation), through ethnic awareness, to ethnic prefer-
ence and then, for some, ethnic prejudice. These illuminate important
reference points for future developmental work in this area and they are
persuasive that mid to late childhood is a major period for the comprehen-
sion and internalisation of many social identities. But, as becomes very clear
through the various discussions, social identities are diverse; development of
any one is multidetermined and intersects with many other powerful social
processes.
Barrett et al., for example, articulate a very clear strategy for how we might
examine SIT/SCT predictions in developmental terms with specific reference
to national identity and they test them against findings from an impressive
multisite research programme. First, they propose we should ask: Do the full
set of specified processes occur at a given age? If they do, then this would
be good news for the generality of the theory. If they do not, this does not
necessarily refute the theory but it leads to interesting further questions: Are
the processes not yet operative or are they there in some nascent form but
overridden by other processes? Barrett et al. note that more complex out-
comes are possible, such as that some of the predicted processes occur and
some do not in children of a particular age. Just to complicate matters a little
further, the authors would doubtless agree that “age” is a rough proxy for
developmental status – so it is possible that children of a given age may be
responding in different ways because of individual differences in the pace of
development.
But if these a priori speculations seem elaborate, the data themselves reveal
a still more complex story. In brief, Barrett et al.’s findings show that different
processes do and do not occur in different age groups in different ways in
different societies, and different patterns are reflected with different meas-
urements. Imagine, worse still, that Barrett et al. were able to report test–
retest data, or to examine longitudinal progress (cf. Rutland, who reports
different within-participant patterns at different times), or to tease out cohort
effects, or to assess international variation in parent–child correspondence
across the several measures employed.
It will be hard for SIT/SCT theorists to ignore the inconvenient data
that Barrett et al. summarise, and those likely to be accumulated as
this kind of work proceeds, on the grounds that the participants are “only
children”. Are social psychological theories to be middle range and age
 Durkin 317
specific? Indeed, in light of Barrett et al.’s findings, we have no reason to
suppose that things become miraculously uniform upon adulthood. On the
other hand, it is salutary to reflect that, had these authors’ purpose been to
test a developmental theory in this domain, such an account would now be
facing equally serious challenges. First, it would not even have generated
several of these hypotheses. Second, those that it would generate (for
example, concerning presumed-universal developmental changes) would be
refuted just as firmly.
Perhaps the question becomes not “when do children start behaving, think-
ing, and feeling in terms of a collective self” (Sani & Bennett) but “given that
children behave, think, and feel in terms of collective selves from the outset,
how do their behaviours, thoughts, and feelings change as they become
increasingly skilled in adapting to the behaviours, thoughts, and feelings of
increasingly diverse others”?

COMPETITION BETWEEN DEVELOPMENTAL AND

SOCIAL PROCESSES

A closely related and particularly intriguing theme that arises is the possibil-
ity of competition among developmental and social processes. Note that this
would scarcely come into view if developmental and social psychologists
remained strictly in their own territories. The contributors here show that
very fertile ground is exposed once we straddle the borders. On the one hand,
we have seen that there are cognitive-developmental changes in children’s
abilities to process social information and respond to contextual cues
(Powlishta, Ruble et al., Sani & Bennett). Developmental changes can modify
some primitive social stances, such as absolutist intergroup attitudes (David
et al., Ruble et al.), as well as allowing the possibility of moderating the
expression of biases (Rutland). During mid childhood children can begin
to appreciate that an out-group (such as the opposite gender, a different
ethnic group) is not homogenous, that external characteristics (such as skin
colour) are superficial, and that people are motivated by different beliefs
(Powlishta, Ruble et al., Sani & Bennett).
On the other hand, social factors such as intergroup competition can over-
whelm developmental capacities, resulting in the establishment of crude
prejudices. Some developmental aspects of social identity may be relatively
predictable on at least a rough age basis but others, such as whether or not a
child develops ethnic prejudices, may depend on motivational factors (Nes-
dale), or on “encounters” (Ruble et al.) with the political brutalities of a given
society that are relatively independent of developmental status, or may even
be affected by the scheduling of international events (see Rutland on the
consequences of Euro ’96).
We noted above that developmentalists have a natural tendency to conceive
of capacities and skills improving. As Ruble et al. suggest, change over time
318 Developmental psychology of the social self
may not invariably be irreversible or unidirectional, and it is possible that
people may experience regressions. These themselves may be instigated or
accentuated by interpersonal and intergroup processes. This collection pro-
vides not only examples but also means of investigating them and, ultimately,
of explaining the intersecting developmental and social phenomena. Nesdale,
for example, illustrates how a theory can be developmental while none-
theless accommodating to the realities of variability in social context.
Rutland describes another approach, by focusing on children’s ability to
adjust to normative beliefs and practices concerning the expression of social
prejudices.
There are points where developmental and social accounts appear to offer
directly competing predictions. A striking example is David et al.’s proposal,
from the perspective of SCT, that during adolescence “gender loses its
prepotency”. On first sight, this is difficult to reconcile with developmental
evidence (1) that gender roles, especially for girls, become more stringently
curtailed during adolescence (Archer, 1984), (2) that adolescents can be more
traditional and more punitive in judging gender-role transgressions than are
primary school age children (Stoddart & Turiel, 1985), and the fact (3) that
sexual attraction and the dictates of entry to the sexual marketplace have
great potency in encouraging adolescents of both genders to align their
appearance and behaviour with contemporary stereotypes of attractiveness
(Basow & Rubin, 1999; Furman, 2002). For many developmentalists, it fol-
lows that adolescence would be a period of intensifying gender-role develop-
ment (Crouter, Manke, & McHale, 1995; Galambos, Almeida, & Peterson,
1990; Huston & Alvarez, 1990).
However, it is true that the cognitive advances of adolescence can, in prin-
ciple, facilitate the capacity to entertain alternative perspectives. And it is true
that, as David et al. note, the social contexts of adolescents’ lives are broaden-
ing. For the SCT theorist, it follows that the fluidity and context variability of
gender categories should increase during this period.
These are apparently competing accounts – and yet they could both be
correct. To offer a simplistic resolution (cautiously, in light of the many
demonstrations in this book that simplistic explanations do not suffice in
accounting for the development of the social self), one possibility is that
increasing cognitive flexibility and context sensitivity at the level of categor-
isation processes jars with increasing community pressure at the level of self-
presentational, interpersonal, and role demands. Part of the developmental
tasks of adolescence, then, is the challenge of “fitting in” (including monitor-
ing of matters of what SCT theorists call normative and comparative fit)
while adjusting simultaneously to the expanded awareness of diversity and
opportunity in the adult world. The competing developmental and social
processes here could lead to intra- and interpersonal conflicts that may not
make gender-role development in adolescence a wholly benign experience;
there is some evidence that gender intensification may be associated with
depression in adolescent females (Wichstrom, 1999).
 Durkin 319
LANGUAGE

Language is the primary social medium through which we transmit informa-

tion and values about the social categories shared within our communities.
Mainstream social psychology’s resilient indifference to language is one of
the continuing mysteries of science, but the assemblage here of several devel-
opmental perspectives brings it repeatedly to our attention. Although none
of the present contributors is primarily concerned with the roles of language
in the development of the social self, many of them touch on interesting
aspects and together the collection suggests that this will be a rewarding focus
for future research.
From a cognitive or cognitive-developmental point of view, the first step
towards a self-concept is the categorisation process – discovering that one is a
member of a particular group (see Sani & Bennett). However, language both
precedes and subserves this process. As Sani and Bennett observe, the main-
tenance of the “we” is not an individual conceptual activity. Through lan-
guage, the categories are already there: One does not create the self label
“Italian” (or “girl” or “black”), one seizes the opportunities to share some-
thing that the culture has already crystallised, identified as important, and
transmitted. Through language, social categories are connected to conceptual
structures (e.g., hierarchies of categories, associations, antonyms) and to
evaluative connotations: One does not decide alone that being Italian is a
good thing; one participates in collective discourse with many others already
convinced of the fact.
The relevance of language to the social self is highlighted here in numerous
ways. One of the fundamental questions is: When is verbal labelling of social
categories possible (Ruble et al.)? How and when do children use verbal
self-labelling? Ruble et al. point to research showing that this is not a straight-
forward age-related phenomenon: Substantially different proportions of
Euro-American and African-American children of the same age (3 years)
self-labelled based on race. Children aged 2½ years rated a person more
negatively when labelled as Arab rather than Israeli. Children acquire social
category labels quite early but simultaneously they attune to the values of
their community.
Recognition of the verbal label for a social identity is an important “first
step” (Sani & Bennett), but it is also a formative one with enduring con-
sequences, possibly (as Sani & Bennett note, and see also David et al., Ruble
et al.) directing attention and providing an anchor. Language may influence
what becomes central and salient (Ruble et al.) and guide the child to the
early stages of social category awareness (Nesdale). The development of lan-
guage is intricately related to the development of knowledge (Ruble et al.). It
is also possible for a label to be used prior to “full” understanding of its
denotation. For example, Barrett et al. report that children sometimes express
strong affective responses to out-group labels well in advance of accumulat-
ing basic factual information about the groups to whom the labels apply. Is
320 Developmental psychology of the social self
this “mere” reference to public criteria (Sani & Bennett), or a developmental
social process that should repay more extensive investigation? It could be that
a framework for categorising and evaluating part of the world is shared and
that subjective identification (Sani & Bennett) emerges as the full meaning is
constructed subsequently through interaction, lexical organisation, and
guided observation.
Several contributors draw attention to the fact that labels for individuals
who transgress stereotypical social category boundaries can be pejorative and
severe. David et al., Powlishta, and Ruble et al. remind us that name-calling
and related verbal derogation are a ubiquitous feature of the practice of
children’s gender segregation. Verkuyten reports that name-calling where the
target is a member of an ethnic minority group is seen by children as the
prototypical instantiation of prejudice and is commonplace in everyday
interaction in some communities. Importantly, Verkuyten shows that this is
not just unpleasant banter but is predictive of the self-esteem of the targets; it
is reasonable to infer that this leads in turn to other social psychological and
affective consequences. Verkuyten’s work also suggests that the language in
which a bicultural or immigrant child is tested can affect orientations towards
personal and social identities.
Through their language, other people convey information about their
categorisation and stereotypes of us. For example, adults modify their
speech style according to the age, gender, or ethnicity of the child, thereby
communicating where they see the individual as slotting into the social
fabric and what capacities and characteristics they assume she or he has (or
lacks). Children, in their turn, are able early on to place someone in a social
category on the basis of the language they speak (Ruble et al.) and are
sensitive to the status of different languages or to language conflicts in their
society (Barrett et al.; Nesdale, 2001). Other people convey, by something as
seemingly minor as personal pronoun choice, whether they see us as part of
the “we” or decidedly aligned with “them” (Banker et al., Ruble et al., Sani &
Bennett).
Language is the foremost means through which we get under each other’s
skin. At present, we lack a comprehensive account of how language relates to
the ontogenesis of social categories and the social self. What these various
contributions make clear is that language is integral and one way in which
developmental perspectives can enrich social psychological work in this area
will be by insisting on closer attention to the medium through which social
identities are encapsulated and shared.

AGE AS A SOCIAL IDENTITY

Ruble et al. make the interesting point that “age” stands out as a likely salient
social identity for children yet has received scant research attention. It seems
again that developmental psychologists, by neglecting social identity, have
 Durkin 321
overlooked an important feature of children’s social and cognitive worlds,
while social psychologists, by studying only one age group, have missed a
social category that may actually be invoked and highly pertinent in many
social contexts. Modern societies are highly age-stratified. Childhood and
adolescence account for about a quarter of the lifespan and within these
phases there are numerous finer gradations, including school levels and more
informal age classifications. Age is very relevant to the social self and is
somewhat unique, in that advances are socially celebrated once a year (and
reviewed every time you meet your relatives): This social category is imbued
with “value and emotional significance” (Tajfel, 1981; see Bennett & Sani).
Although this was not their principal focus, Sani and Bennett’s work sup-
ports the inference that age is integral to the developing social self: Children
describe themselves in a different way when primed with the concept of an
adult. Similarly, David et al. report lower gender stereotyping when the con-
trast is with “grown-ups” rather than gender. Closely correlated with this is
the fact that age status is very important to the people with whom children
interact: As Powlishta points out, adults make stereotypical inferences about
children as children.
SCT in particular seems to offer an explanation of the seeming paradox
that young children betray strongly negative social stereotypes about elderly
people (Davidson, Cameron, & Jergovic, 1995; Goldman & Goldman, 1981)
yet demonstrate great affection (positive in-group bias) when asked about
their grandparents and great-grandparents (Kaiser, 1996). Although age may
be an important social category it seems that it, too, is not a fixed construct
but is context-sensitive.

SOCIAL IDENTITIES ARE DIVERSE – AND MAY

DEVELOP DIFFERENTLY

Another reason why this book cannot offer a global template for develop-
mental stages in the emergence of the social self is the accumulating evidence
it provides that “social identities differ” (Sani & Bennett) and hence are
acquired differently. Social identities differ in content and complexity (e.g.,
external appearance vs. belief system), in terms of personal salience (Ruble
et al.), and in terms of how they fit into the broader society (e.g., virtually
everyone acquires a gender identity of some sort, but not everyone has a
religious identity, and though most people have an ethnic identity it is likely
to have different salience according to whether or not one is in a minority
group and, after Verkuyten, which minority group it is).
Ruble et al. report valuable new findings on the parallel progression in the
development of gender and ethnic understanding. As they argue, this sup-
ports the assertion of a central role for cognition, though we cannot entirely
rule out a contributory role for comparable task demands. Nevertheless, the
authors stress that social context can affect the developmental processes and
322 Developmental psychology of the social self
sometimes may offer better accounts of the data. They describe differences
among children of different ethnic backgrounds on developmental tasks that
are both poignant and theoretically attention-grabbing.
Rutland makes a convincing case that different forms of prejudice are
subject to different levels of societal tolerance. Ethnic prejudice is at least
controversial and often, if not invariably, condemned, whereas national
prejudice is at least ambivalent: In most countries, it is acceptable to sub-
scribe to the view that “our nation is the finest on Earth” and to express
hostile views to at least some outsiders. Gender prejudice is different again: it
is normative in some cultures, implicit in others, confronted by contemporary
ideology in some, and perhaps occasionally acceptable in polite company
when couched in antitraditional terms (e.g., directed at males).
The greater in-group gender bias among girls, noted in several chapters
(David et al., Ruble et al., Powlishta, Rutland), is a further empirical reason
to avoid the assumption that the development of social identities is uniform.
Even in the same domain, it appears that different groups develop differently.
It could be that the differential in-group bias is simply factually underwritten:
Maybe boys are more troublesome, making it relatively easy for girls to derive
self-esteem from in-group identification. It could be that it is socially transmit-
ted: Parents and teachers are prone to say that boys are more troublesome. It
could be that it is motivated by girls’ sensitivity to social inequities (David et
al., Powlishta). It could be that both genders are sensitive to prevailing norms
and are operating as “social tacticians” (Rutland); for example, the norms of
traditional paternalism and political correctness leave it less acceptable to
criticise females but tolerable to denigrate males. Whatever the explanation, it
does appear that children are sensitive to the value systems of their societies
(Tajfel, 1981) and that this sensitivity bears on the development of social
selves, resulting in different experiences and, arguably, different routes.

CAN WE CREATE BETTER SOCIAL SELVES?

The nature of the subject matter means that every chapter in this book deals
with real-world phenomena of pervasive significance. Social identities matter:
They are directly related to social status, to equity, to opportunity, to self-
esteem, and to well-being. A natural consequence, not far from the surface in
most chapters and addressed directly in some, is the question of whether
developmental- and social-psychological research can be applied to promote
more favourable outcomes in terms of how developing people adapt to the
realities of social categories and how they treat members of out-groups.
The heartening news may be that the “process of social categorisation . . .
is not completely unalterable” (Banker et al.). If the potentially volatile arena
of stepfamily relations over time is less conflict-ridden when the merged ini-
tial families come to think of themselves as one group, then the scope for
decategorising and recategorising (Banker et al.) becomes of wide interest.
 Durkin 323
In a similar vein, Nesdale’s work also suggests grounds for optimism. In his
studies, children certainly show in-group bias but this is not necessarily iden-
tical with in-group ethnic bias. If different ethnicity peers are included in the
in-group, they are liked no less than same-ethnicity peers.
Modifying social categories and social identities is not a simple task and
none of the contributors underestimates the magnitude, complexity, and per-
haps dangers of such intervention. While there may be grounds for optimism,
there are also good reasons for caution. After all, social categories didn’t get
where they are today by being ephemeral. There is an edge to all the sugges-
tions and demonstrations of intervention: The risk is ever-present that the
underlying processes that set the groups apart in the first place could be
re-invoked suddenly and powerfully, exacerbated by a strong infusion of
reactance. As the present authors remind us repeatedly, much depends on the
context.
Rutland’s research shows that the expression of intergroup prejudice may
be ameliorated by (developmental changes in) processes of normative
behaviour. His clever use of a dissociation paradigm highlights an important
qualification and a challenge to future work, namely that children may
become adept in achieving socially desirable self-presentations without neces-
sarily relinquishing automatic biases. As Rutland points out, it is good to
know what not to do – but this still presupposes knowledge of what could be
done.
Powlishta ventures the bold proposal that better gender relations can be
fostered by encouraging children to self-categorise on a dimension other than
gender. David et al. disagree, arguing that self-categorisations are “not a
matter of choice but a direct consequence of the genuine economic, political,
and social circumstances of our lives”. Powlishta might reasonably respond
that someone chooses the political and social circumstances, and choices are
open to review. Each perspective captures important dimensions and together
they remind us that social change is inherently dialectical. Psychologists
interested in these issues may have to decide whether or not it is their
responsibility to subscribe to that overarching social category, the status
quo.

SOCIAL SELVES THROUGH THE LIFESPAN

I will be brief here because this is very much an area that this book designates
for future research rather than tackles directly. With the exception of Banker
et al., the contributions have been concerned with child development. There is
plenty to say about child development and the contributors have said much;
we can rely on our colleagues in social psychology to say at least as much
about the social identities and self-categorisations of 18- to 21-year-olds.
Together, these provide a fitting starting point for lifespan research, and most
of that lies ahead – in studies of the next three quarters of life.
324 Developmental psychology of the social self
We know relatively little of how people’s orientations towards social iden-
tities and their processes of self-categorisation change as they progress
through life. The overwhelming expectation we can derive from this book is
that we will need to draw on both developmental and social-psychological
perspectives if we are to investigate adequately the longer-term construction
and maintenance of social selves.
The contributions and issues arising in this volume do suggest many tantal-
ising avenues for inquiry. Several contributors note that the phenomena they
are investigating are by no means unique to childhood. Ethnic prejudice
obviously does not necessarily wane with chronological maturity (Nesdale).
National identity is of increasing interest during childhood and adolescence
(Barrett et al., Sani & Bennett), and is likely to remain prominent henceforth
for many, though its centrality and salience may well be affected by both
developmental/chronological attainments (such as being granted adult citi-
zenship rights, falling within the conscription age-span) and the social impact
of life events (such as war and terrorism).
Again, families are often the longest-lasting groups of people’s lives,
though the ways in which people relate to them – represent them as part of
the social self – are likely to vary with lifestage and responsibilities. Recall the
implication of Verkuyten’s evidence that, among the people most disadvan-
taged by prejudice, a person’s developmental status may well influence how
she or he responds to it (e.g., parents may try to bolster their children).
Banker et al. study stepfamily conflict largely at the level of the unit, but
implicit in their account is the important fact that individuals have different
roles and expectations as a function of developmental status. Similarly, an
important part of the social self for most people is their occupational role:
This changes through the lifespan. Occupational identities are not constant
(indeed, at some point they cease) and they interact with other aspects of the
social self that also vary with age and lifestage, not least gender identity (see
David et al., Powlishta, for discussions of the persistence of gender prejudice
into adulthood). It would be strange indeed if social selves did not develop
through adulthood; stranger still if we ignored the fact.

CONCLUSIONS

I greatly enjoyed reading this collection, as, I imagine, will many readers. I
also imagine that readers will find it difficult, as I did. The difficulty lies not in
any of the individual chapters, each of which is lucid, well-structured, and
well-documented, even though they do tackle very complex issues. It lies in
the ambition of the collective venture: the integration of developmental and
social psychological perspectives on the social self. Developmental and social
psychologists, as the editors eloquently summarised in the opening chapter,
may be closely related but they rarely speak the same language. They are
differentiated by whole bodies of assumptions, vocabularies, methods, per-
 Durkin 325
spectives, and even explanatory goals. It is very difficult to reconcile them but
the bilingual skills represented in this volume demonstrate that it is very
worth while to attempt to do so.
The book raises many new questions and promises to inspire a vast array
of new research. But it does not just set puzzles: It assembles a large amount
of creatively won, thoughtfully interpreted empirical evidence and it estab-
lishes some important guidelines for future work. In particular, it has made
major headway in casting the (already, as the editors comment, colossal)
framework of social identity theory and self-categorisation theory into sharp
developmental relief. It will also compel developmentalists to reconsider their
conception of stereotypes not as a fixed body of knowledge but as “inher-
ently comparative, flexible, and variable” phenomena (Sani & Bennett). It
may persuade social psychologists to come to terms with the developmental
phenomena that are integral to their participants’ social being. Above all, it
confirms that developmental science becomes richer when it merges the
laboratory with the real world.

REFERENCES

Archer, J. (1984). Gender roles as developmental pathways. British Journal of Social

Psychology, 23, 245–256.
Basow, S. A., & Rubin, L. R. (1999). Gender influences on adolescent development. In
N. G. Johnson, M. C. Roberts, & J. Worrell (Eds.), Beyond appearance: A new look at
adolescent girls (pp. 25–52). Washington, DC: American Psychological Association.
Battistelli, P., & Farneti, A. (1991). Grandchildren’s images of their grandparents:
A psychodynamic perspective. In P. K. Smith, (Ed.), The psychology of
grandparenthood: An international perspective (pp. 143–156). London: Routledge.
Bowlby, J. (1988). A secure base: Parent–child attachment and healthy human
development. New York: Basic Books.
Crouter, A. C., Manke, B. A., & McHale, S. M. (1995). The family context of gender
intensification in early adolescence. Child Development, 66, 317–329.
Davidson, D., Cameron, P., & Jergovic, D. (1995). The effect of children’s stereotypes
on their memory for elderly individuals. Merrill-Palmer Quarterly, 41, 70–90.
Furman, W. (2002). The emerging field of adolescent romantic relationships. Current
Directions in Psychological Science, 11, 177–180.
Galambos, N. L., Almeida, D. M., & Petersen, A. C. (1990). Masculinity, femininity,
and sex role attitudes in early adolescence: Exploring gender intensification. Child
Development, 61, 1905–1914.
Goldman, R. J., & Goldman, J. D. G. (1981). How children view old people and
ageing: A developmental study of children in four countries. Australian Journal of
Psychology, 3, 405–418.
Huston, A. C., & Alvarez, M. M. (1990). The socialization context of gender role
development in early adolescence. In R. Montemayor, G. R. Adams, & T. P.
Gullotta (Eds.), From childhood to adolescence: A transitional period? Advances in
adolescent development: An annual book series, Vol. 2 (pp. 156–179). Beverly Hills,
CA: SAGE Publications.
326 Developmental psychology of the social self
Kaiser, P. (1996). Relationships in the extended family and diverse family forms. In
E. E. Auhagen, & M. von Salisch (Eds.), The diversity of human relationships
(pp. 141–170). Cambridge: Cambridge University Press.
Nesdale, D. (2001). Language and the development of children’s ethnic prejudice.
Journal of Language and Social Psychology, 20, 90–110.
Stoddart, T., & Turiel, E. (1985). Children’s concepts of cross-gender activities. Child
Development, 56, 1241–1252.
Tajfel, H. (1981). Human groups and social categories: Studies in social psychology.
Cambridge: Cambridge University Press.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorization theory. Oxford: Blackwell
Publishing.
Wichstrom, L. (1999). The emergence of gender difference in depressed mood
during adolescence: The role of intensified gender socialization. Developmental
Psychology, 35, 232–245.
Author index

Abecassis, M. 300, 303 Arbuckle, J.L. 279, 280 Bartel, N.R. 223
Abell, J. 86, 114, 164, 177 Archer, J. 318 Bartsch, K. 78
Aberson, C.L. 117, 295 Arcuri, L. 255, 270 Basow, S.A. 318
Aboud, F.E. 20, 29, 33, Ardelt, M. 295 Bauer, P.J. 107, 109–110
35–37, 38, 44, 47, 52, 54, Arnold-Dorman, C. 255 Baumeister, R.F. 53, 196,
56, 61, 159, 191, 207, 219, Aronson, J. 59 293
220, 221, 222, 223, 224, Ashburn-Nardo, L. 296 Beal, D.J. 270
227, 228, 229, 230, 233, Asher, S.R. 223, 226 Beaudet, J. 223, 233
248, 252 Aube, J. 55, Beeghly-Smithy, J. 33
Abrams, D. 2–3, 6, 21, 51, 53, Aubry, S. 56, 58 Beere, C.A. 110
60, 91, 225, 248, 253, 258, Augoustinos, M. 248 Bem, S. 39, 43, 116, 140, 146,
269, 294, 295, 296, 297, Averhart, C. 49 299
298, 299, 301, 302, 304, Ashmore, R.D. 30, 46 Benet-Martinez, V. 206–207
305, 306, 314 Bennett, M. 17–18, 41,
Adorno, T.W. 220 Bachman, M. 2, 17, 91–92, 43–44, 62, 79, 80, 87–88,
Ageton, S.S. 300 267, 268, 269, 272, 275, 92–93, 96, 108, 110, 115,
Agostini, L. 43, 79, 80, 108, 314–318, 320–322 121, 124, 160, 165, 175,
121 Bacon, S. 295 177, 181, 192, 248, 251,
Akiba, D. 34, 45, 47, 59 Bacue, A. 138 254, 258, 293, 295–297,
Alarcon O. 47, 54 Bagley, C. 198 302, 306, 314–317,
Albert, A.A. 52, 110 Bahrick, L.E. 135 319–321, 324–325
Alejandro-Wright, M.N. 39 Bailey, R.C. 229 Berger, P. 297
Alfieri, T. 292 Balaam, B.J. 256 Berglas, S. 4
Allen, V.L. 51, 223, 226 Baldwin, J.M. 1–2 Bernal, M.E. 36, 37, 159
Allen, W.R. 49 Banaji, M.R. 135, 206, 255, Berndt, T.J. 110
Allport, G.W. 4, 106, 221, 258 Berger, S.H. 277
267, 268, 269, 271, 272, Bandura, A. 46, 58, 138–142 Best, D.L. 86, 110, 137, 223
274, 279 Banerjee, R. 90, 92–93, 96, Bhatt, S. 253
Almeida, D.M. 318 254 Biernat, M. 60, 110
Alvarez, J.M. 2, 17, 29, 33, Banker, B. 21, 268, 276, 279, Bigler, R.S. 3, 31, 38, 43, 44,
44, 46, 47, 52, 56, 63, 283, 305, 315, 320, 322–324 49, 50, 62, 104–106,
91–92, 192, 197–198, 220, Banks, W.C. 53, 191, 223 114–115, 117–118, 122,
248, 291, 294, 314–318, Barájas, L. 205 126, 137, 162, 182, 192,
320–322 Bargh, J.A. 255 222, 223, 234, 248, 250, 251
Alvarez, M.M. 318 Baron, R.M. 278 Billig, M. 7–8, 104, 147, 225,
Amato, M. 35, 52, 248 Barreto, M. 253 253
Anastasiades, P. 257 Barrett, M. 19, 62, 78, 86–87, Bird, C. 221
Anastasio, P.A. 267, 268, 272, 90, 124, 159, 160, 161, 165, Bittinger, K. 107
275 166, 168, 175, 176, 177, Black-Gutman, D. 233
Anderson, D.R. 56 178, 180, 181, 192, 248, Blaine, B. 50–51,
Anderson, J.Z. 272, 277 293, 302–303, 314, Blakemore, J.E.O. 138
Andrews, G. 94 316–317, 319–320, 324 Blaske, D.M. 109
Annis, R.C. 52, 194, 209 Bar-Tal, D. 62, 77 Bodenhausen, G.V. 253, 259
Anthony, E.J. 268 Bartel, H.W. 223 Bolduc, D. 137
328 Author index
Bolger, N. 33, 46, 201, 294 Buss, A.A. 3 Collman, P. 109
Bond, M.C. 195 Bussey, K. 46, 58, 135, Collins, P.A. 56
Booth, A. 268 138–142 Condor, S. 301
Boswell, D.A. 223 Byrne, D.F. 231 Condry, J. 136
Boulton, M.J. 204, 229, 233 Condry, S. 136
Bourchier, A. 161, Cable, I. 164 Conner, J.M. 254
Bourhis, R.Y. 162 Cairns, E. 81 Connolly, P. 34, 96–97
Bovill, M. 299 Caldera, Y.M. 137 Cook, K.V. 137, 279
Bowers, C. 192 Cameron, L. 298 Cooke, T. 94
Bowlby, J. 29, 314 Cameron, J.A. 2–3, 17, 33, 45, Cooley, C.H. 1–2, 29
Bown, N. 60, 301 52, 61–62, 91–92, 197–198, Cooper, J. 59
Bradbard, M. 56 220, 248, 291, 314–318, Cooper, M.J. 256, 257
Brand, E.S. 35, 61, 191, 220, 320–322 Cordua, G.D. 111
224 Cameron J.E. 42, 57 Corenblum, B. 52, 194, 209
Branch, C.W. 232 Cameron, L. 258 Cornell, S. 194
Brandstadter, J. 4 Cameron, P. 321 Cossette, L. 137
Brannon, L. 140, 145 Campbell, A. 299 Cota, M.K 37
Branscombe, N.R. 51, 54, 92, Campbell, J. 160, 248, 249 Cotting, D. 3, 85, 296
114, 116, 162, 164, 180, Cann, A. 110 Coughlin, R. 223
194, 232, 253 Canter, R.J. 300 Cousins, S.D. 200
Brauer, M. 298 Cantor, N. 3 Cowan, G. 110
Bray, J.H. 277 Capozza, D. 251 Coyne, M.J. 109
Brazil, D.M. 268 Carter, D.B. 43, 109, 116 Craig, J. 55
Brehm, S.S. 4 Carver, C.S. 258 Cranwell, F.R. 200, 205
Breinlinger, S. 149 Case, R. 3 Crnic, K. 189, 207
Brendgen, M. 299, 299 Cash, T.F. 256 Crocker, J. 42, 45, 49–51,
Bretherton, B. 33, Caspi, A. 295 53–54, 195, 201
Brewer, M.B. 3, 29, 31, 51, Castelli, L. 255 Crook, C. 299
61–62, 108, 193, 198, 225, Catalyst 138 Crosbie-Burnett, M. 272, 277
272, 274 Cecil, C.L. 268 Cross, S.E. 143, 296
Broadnax, S. 50–51, Celious, A. 209 Cross, W.E. 31–2, 39, 53, 59,
Bromley, D. 29, 33, 45, 106 Chaiken, S. 255 191, 194, 196
Bronfenbrenner, U. 192, 207, Chan, S.Q. 39 Crouter, A.C. 318
292 Charlesworth, W.T. 118 Czopp, A.M. 296
Brookes, J 135 Charters, W.W. 114
Brookes-Gunn, J. 3 Chen, L.49, 55 Damon, W. 3, 29, 45, 78, 111
Brooklins, G.K. 49 Cheyne, W.N. 249 Davey, A. 43, 44, 61, 86, 192,
Brotherton, C.J. 114 Chavira, V. 32, 221, 222, 223, 225, 226, 232
Brown, A.C. 277 Chavous, T.M. 42, 46, 49, David, B. 18, 20, 91, 118, 120,
Brown, B.B. 136, 162, 50–51, 53, 194, 195 138, 140, 143, 145, 292,
Brown, G. 227 Chi, M.T.H. 85 294, 296, 298, 299–301,
Brown, K. 237 Chiu, C.Y. 206–207 302, 318–324
Brown, L.M. 195 Child, P. 44, 293 Davidson, D. 321
Brown, R.J. 31, 43–44, 50–51, Chodorow, N. 139 Davis, H. 221,
61, 104–107, 112, 118–119, Chyatte, C. 232 Davis, S.W. 114
121, 162, 164, 192, 193, Citron, C.C. 254 Day, J. 177
207, 221, 222, 225, 229, Clack, F.L. 195, 199 Deaux, K. 3, 53, 85, 114, 296,
232, 233, 234, 239, 248, Clark, A. 231 297
249, 251, 252, 270, 274, Clark, K.B. 53, 190, 227 Delk, J.L. 137
298, 305 Clark, M.P. 53, 190, 227 Del Miglio, C. 292
Brucken, L. 109 Clemens, L.P. 35, 107, 109 del Valle, A. 19, 90, 293,
Bruner, J.S. 13 Cloud, M.J. 86 302–303, 314, 316–317,
Bryant, W.T. 112, 118, 122 Coard, B. 198 319–320, 324
Buehrig, C.J. 229 Coats, S. 195 Dembo, M.H. 231
Bukowski, W.M. 299 Coker, D.R. 44, 58 Demetriou, A. 4
Bundy, R.F. 7–8, 104, 147, Colburne, K.A. 109, 113, 120 Demo, D.H. 295
225 Coleman, H.L.K. 205 Denmark, F. 114, 123
Burchardt, C.J. 254 Coleman, J. 51, Deschamps, J.C. 112,
Burdick, H. 221 Coll, C.G. 2, 17, 91–92, 189, 114–115, 120
Buriel, R. 39, 42 207 Devine, P.60, 255, 296, 305
 Author index 329
DeVos, G.A. 194 Fabes, R.A 48, 58, 125 Gardner, W. 29
De Vries, M. 198 Fagan, J.F. 107 Garfinkle, G.S. 45
Diaz, I. 135 Fagot, B.I. 35, 56, 107, 109, Garza, C.A. 37, 195, 209
DiMartino, L. 45, 47, 59 135–137 Gelman, S.A. 31, 109, 111,
DiMiceli, A.J. 114, 123 Fairburn, C.G. 256, 257 118
Dinella, L.M. 57 Fairlie, P. 199 Genesee, F. 229
Dion, K.L. 120 Farris, K.R. 205 George, D.M. 223
Dobliner, D.B. 192 Farroni, T. 161 Gergen, K.J. 6
Doise, W. 6, 9, 105, 112, Feagin, J.R. 49 Gerton, J. 205
114–115, 120 Feffer, M. 231 Gibbons, F.X. 4
Donnelly, B.W. 277 Fein, S. 51, 54 Gibson, W.J. 92
Doosje, B. 92, 114, 144, 164, Feinman, S. 137 Giles, H. 86, 122
180, 251 Fenigstein, A. 3 Glass, C. 253
Dovidio, J. 21, 51, 255, 267, Ferguson, C.K. 7, 79–80 Glick, P. 112, 118, 122–124,
268, 269, 271, 272, 273, Ferrier, G. 93 298, 301
275, 276, 278, 279, 283, Fhagen-Smith, P. 32 Gochberg, B. 55
294, 305, 315, 320, 322–324 Fields, J.P. 54 Goldberg, P. 123
Doyle, A.B. 52, 112, 118, Fine, G.A. 207 Goldman, J.D.J. 321
121–122, 136–137, 192, Fine, M. 192, 277 Goldman, R.J. 321
221, 222, 223, 233, 248, Finkel, E.J. 268, 283 Good, C. 59
249, 252, 254, 256, 259, 299 Fischer, K.W. 3 Goodman, M.E. 190, 221,
Drabman, R.S. 111 Fishbein, H.D. 191, 198, 224, 224, 226, 227, 232
Druickman, J. 277 229, 248 Goodnow, J. 33
Duncan, B.L. 51, Fishman, B. 284 Goto, S.G. 206, 305
Durkin, K. 4–5, 21–22, 29, Fiske, S. 43, 85, 135, 272, Gough, H. 221
222, 234, 235, 248 298 Gourevitch, V. 231
Dustin, D.S. 114 Flament, C. 7–8, 104, 147, Govender, R. 255
Duval, S. 259 225 Grace, D. 18, 20, 91, 118, 120,
Dweck, C. 33, 59 Flavell, J.H. 4, 222, 230 140, 292, 299–301, 318–324
Flesser, D. 49, 105, 162, 182, Gracely, E. 35, 107, 109
Eagly, A.H. 50, 61, 149 225, 234, 239, 248, 251 Gralinski, H. 36
Eckes, T. 4, 6, 159 Flett, G.L. 86 Gramzow, R.H. 270
Eckles, J. 57, 298 Ford, L. 3 Granrose, C.S. 205
Edelbrock, C. 109 Fordham, C. 59 Grant, G. 192, 223, 229
Eder, R.A. 33, 78, 299 Fowles, J. 86 Grant, P.R. 162
Edwards, C.P. 86 Frable, D.E.S. 209 Green, R.J. 277
Egan, S.K. 55, 118, 122 Frazier, S. 274 Greenwald, A.G. 206, 255,
Ehrhardt, A. 135 Frenkel-Brunswik, E. 220, 257, 258
Eichstedt, J.A. 109, 110, 113, 221 Greenwald, H.J. 226
118, 122, 124, 298 Freud, S. 139 Gregor, A.J.220,
Eisenbud, L. 108 Frey, K. 87, 31, 33, 56, 59 Greulich, F. 55, 87
Eisenberg, J. 199 Fujioka, T. 44, 293 Grifiths, J. 234, 235
Eiser, J.R. 9 Fuligni, A. 2, 17, 33, 45, Grill, J.J. 223
Eisner, H. 208 52–53, 91–92, 197–198, Guinote, A. 298
Ellemers, N. 144, 164, 180, 220, 248, 291 Gulko, J. 58, 106, 109–110,
195, 251, 253 Furman, W. 318 112–113, 119–120, 122
Elman, M. 118 Furstenberg, F.F.Jr. 268, 277 Gurtman, M.B. 51, 271, 294
Elmian, M. 136, 137 Fyffe, C. 250 Gutek, B.A. 138
Elms, A.C. 6
Emler, N. 192, 199, 207, 294 Gaertner, S. 21, 267, 268, 269, Haaf, R.A. 109
Endsley, R. 56 270, 272, 274, 275, 276, Hafan, R. 137
Epperson, S.E. 148 278, 279, 283, 295, 297, Hagendoorn, L. 196, 208
Erikson, E. 31 305, 315, 320, 322–24 Hains, S.C. 270
Erkut, S. 47, 54 Gaines, S.O. 205 Hall, S.K. 55,
Etaugh, C. 110 Galambos, N.L. 318 Halpern, D.F. 148
Etcoff, N.L. 43 Gallagher, J.J. 224 Halverson, C.F. 43, 56, 84,
Ethier, K.A. 53 Garciá, B.F. 205 106, 111, 116, 121
Ethridge, R. 35 Garcia Coll, C. 34, 45, 47, 49, Hamilton, D.L. 85, 164
Ethridge, T. 107, 109 54–55, 59, 291 Hamm, M.S. 232
Evans, S.M. 58 Garcia, H.V 189, 207 Harris, A.M. 50
330 Author index
Harris, D. 221 Hong, Y.Y. 206–207 Johnson, S.P. 227
Harris, J.R. 135, 142 Hoover, R. 248 Johnston, E. 107
Harris, P. 94 Hopkins, N. 43, 86, 108, 114, Jones, E.E. 3–4
Harrison, A.O. 39, 121, 164, 177 Jones, J.M. 29, 44, 53, 56
Hart, D. 3, 29, 45, 78 Hoppe, R.A. 223 Jones, L.C. 31, 50, 105–106,
Hartman, D. 194 Hornsey, M.J. 305 114, 117, 162, 182, 192,
Harter, S. 2–4, 29, 33, 87, 117, Hort, B.E. 109 234, 248, 250
142, 196 Horwitz, M. 7 Jopling, D. 142
Hartup, W.W. 300, 303 Horowitz, E.L. 190, 221, 222, Jost, J.T. 293, 297, 304
Harvey, R.D. 54, 194 Horowitz, D.L. 194, 221, 222 Joyner, L. 87–88, 177, 302
Haslager, G.J.T. 300 Houghton, S. 248 Judd, C.M. 201, 298
Haslam, S.A. 4, 10–15, 17, Houlette, M. 21, 275, 315, Judge, J. 248
86–87, 91, 93, 114, 162, 320, 322–324 Jussim, L. 30
163, 164, 177, 190, 193, Howard, A. 255
256 Howard, J.M. 270 Kafati, G. 272
Haugh, S.S. 110 Hraba, J. 192, 223, 229 Kaǧitçibaşi, C. 199
Haunschild, P.R. 268 Hu, L.T. 273 Kail, R.V.Jr. 108
Havel, J. 221 Hubbard, J. 299 Kaiser, P. 321
Hawker, D.S.J. 2004 Hughes, D. 49, 55 Kanagawa, C. 296
Hayden-Thomson, L. 112 Huiberts, A.M. 200 Kardiner, A. 191
Hayes, B.K. 86–87, 164, 177 Hunsberger, B. 226 Karpinski, A. 270
Healy, M. 117, 295 Hur, Y.M. 295 Kashima, Y. 293
Heiman, R.J. 195, 199, 205 Hurtig, M.C. 119, 149 Kashy, D.A. 201
Heine, S.H. 294, 304 Huston, A.C. 137, 254, 318 Kassin, S.M. 4
Helbing, N. 109 Hutnik, N. 198 Katz, P.A. 35–36, 38–39, 49,
Helgerson, E. 229 Hutton, D. 94 107, 109, 111–112, 114,
Heller, K.A. 110 Hwang, C.H. 52 118, 122, 136, 220, 224,
Helmreich, R. 55 Hymel, S. 112, 300 227, 228, 230, 248, 252,
Helwig, A.A. 109 301
Henderson, M.C. 205 Iacono, W.G. 295 Kawakami, K. 255
Henson, N. 60, 301 Imai, S. 224, 229 Kazi, S. 4
Henry, S. 270 Insko, C.A. 268 Kee, D.W. 108, 118–119
Heppen, J. 30 Intons-Peterson, M.J. 49 Keefe, S.E. 197
Herringer, L.G. 195, 209 Iuzzini, J. 295, 297 Keller, A. 3
Hetherington, E.M. 295 Kelley, H.H. 7
Hetts, J.J. 205–206 Jacklin, C.N. 103, 112, Kelly, C. 149, 162
Hewstone, M. 31, 51, 54, 60, 136–137 Kenny, D.A. 201, 278
228, 270, 273, 274, 295, Jackson, D.W. 137, 195 Khun, D. 109
305 Jackson, S. 199 Kiesler, S.B. 123
Heyman, G.D. 118, 119 Jahoda, G. 160, 161, 224, 248, Kihlstrom, J.F. 3
Heywood, C.299 249, 250, 253 Kinket, B. 196, 200, 202, 208,
Hicks, D. 205 James, A. 300 250
Hickson, F. 233 James, S.D. 272, 277, 284 Kitayama, S. 142, 195, 199,
Higgins, E.T. 3, 29, 34, 292, James, W. 1–2 205, 294, 304
295 Jansen, V.G. 224 Kite, M.E. 149
Hilt, L. 112, 118, 122–124, Janssens, L. 114 Klein, S. 118
298, 301 Jaspars, J.M.F. 6, 160, 161, Klineberg, O. 78, 86, 160,
Hinkle, S. 162 165 161, 248
Hirschfield, L.A. 41, 135 Jenkins, R. 97, 189, 207 Klink, A. 162
Ho, H. 108 Jenks, C. 300 Knight, G.P. 36–37, 159
Hocevar, D. 231 Jergovic, D. 321 Knonsberg, S. 137
Hoffman, C.D. 110 Jetten, J. 253 Koestner, R. 55
Hofstede, G. 195, 199, 206 John, O.P. 91 Kofkin, J.A. 35–36, 39, 49
Hogg, M.A. 2, 6, 11, 13–14, Johnson, B. 255, 274, 277, Kohlberg, L. 29, 30, 37–38,
16, 31, 51, 53, 60, 79, 91, 284 56, 58, 139, 147, 231
113–114, 123, 163, 190, Johnson, D.W. 272 Koomen, W. 114, 200
220, 225, 228, 269, 270, Johnson, K. 21, 94, 315, 320, Kopp, C. 36,
294, 295, 298, 302, 304, 322–324 Kortekaas, P. 164, 195
305, 315 Johnson, N. 160, 161, 165, Kozar, R. 268
Holmes, R.M. 40, 44 224, 233, 248, 249, 250 Kramer, R.M. 91
 Author index 331
Kruglanski, A. 294 124, 160, 161, 165, 175, McGee, D.E. 255, 257, 258
Ksansnak, K.R. 136 176, 181, 192, 248, 293, McGeorge, P. 258
302–303, 314, 316–317, McGlynn, E.A. 52, 63
LaFreniere, P. 118 319–320, 324 McGraw, K.O. 111
LaFromboise, T. 205 McGue, M. 295
Lalonde, R.N. 42, 196 Maass, A. 234, 235, 255, 270 McGuire, C.V. 44, 293
Lambert, W.E. 78, 86, 160, Maccoby, E. 48, 52, 103, McGuire, S. 295
161, 229, 248 109–110, 112–113, 118, McGuire, W.J. 6, 44, 293
Lamberty, G. 189, 207 120, 122, 123–124, McHale, S.M. 318
Lamborn, S.D. 136 136–137, 141, 249, 298 McKillip, J. 114, 123
Larue, A.A. 138 Mackie, D.M. 104–106 McKenry, P.C. 277
Larwood, L. 138 MacPherson, J. 88 McLaughlin K. 236
Lay, C. 199 Macrae, C.N 253, 259 McNew, S. 33
Leary, M.R. 53, 196 Madden, R.B. 137 McPherson, D.A. 220
Lee, J. 108 Madden, T. 195 Meacham, J. 3
Lee, H.K. 200 Madge, N.J.H. 231 Mead, G.H. 1–2, 29, 297
Lehman, D.R. 294, 304, Madole, K.L. 107 Meeus, W. 200
Leinbach, M.D. 35, 107, 109, Madson, L. 143 Melamud, A. 199
135–137 Magnusson, D. 34, 45 Messick, D.M. 104–106
Lemmon, K. 4 Mahan, J. 192 Meyer, G. 114
Lerner, R.M. 4, 229 Major, B. 53, 54, 114, 195 Meyers, B. 107, 109
Letourneau, K.J. 56 Malcuit, G. 137 Middleton, M. 160, 161, 165,
Leung, K. 195, 199 Malucchi, L. 43, 79–80, 108, 233
Levin, J. 232 121 Miller, L.F. 268
Levine, J.M. 207 Manke, B.A. 318 Miller, N. 193, 272
Levine, R.V. 108 Mann, J.A. 267, 274, 278 Miller, R.E. 96, 107–108
Levinson, D.J. 220 Manstead, A.S.R. 92, 232, Mills, C.J. 119
Levy, G.D. 43, 107,109, 111, 253 Milne, A.B. 253, 258, 259
116, 118 Marcia, J. 31 Milner, D. 191, 192, 198, 222,
Lewis, M. 3, 29, 135–137 Marelich, W.D. 205 223, 224, 225, 226, 227,
Leyens, J.P. 237 Marie-Covatto, A. 294, 300, 228, 231
Liben, L. 43, 62, 118, 126, 301 Mischel, W. 139
137, 223, 248 Markell, M. 62, 162, 248, 251 Mitchell, F.G. 223, 224, 229,
Liebl, M. 110 Markham, E. 267 233
Liebkind, K. 197 Markiewicz, D. 299 Mizrahi, K. 3, 85, 296
Lin, M. 272 Markman, E.M. 31 Mladinic, A. 50, 61
Lintern, V. 90, 253 Markstrom-Adams, C. 52 Moffit, T.E. 295
Linville, P.W. 3 Markus, H. 3–4, 57, 142, 143, Moghaddam, F. 104, 196
Little, J.K. 109, 110 195, 199, 205, 294, 296, 304 Moller, L.C. 113, 120
Liss, M.B. 35, 39, 52, Marques, J.M. 3, 60, 237, Monachesi, E.D. 221
Livesley, W. 29, 33, 45, 106 248, 298, 301 Money, J. 135
Livingston, M. 137 Marsh, A. 227 Monteith, M.J. 296, 299, 305
Livingstone, S. 299 Martin, C.L. 29, 33, 35–37, Montemayor, R. 57
Lobliner, D.B. 3, 31, 50, 62, 40, 43, 48–49, 84, 50–51, Montgomery, M.J. 299, 301
104–106, 114, 117, 162, 55–60, 106, 108–112, 116, Montoya, M.R. 268
182, 234, 248, 250 118, 121, 125, 159 Moore, C. 4
Lockheed, M.E. 50, 118 Martin W.E. 221, Moreland, R.L. 269
Lohaus, A. 109 Martinez, C.M. 232 Morland, J.K. 52, 191, 226,
Long, K.M. 232, 253 Martinez-Taboada, C. 60 232
Luckmann, T. 297 Masser, B. 299, 305 Morris, E. 177
Luebke, J. 114, 123 Masson, K. 200 Morris, M.W. 206–207
Luecke-Aleksa, D. 56, Masten, A.S. 299 Morrison, E. 255
Luhtanen, R. 42, 45, 49–51, Masters, J.C. 4 Moscovici, S. 6, 199, 297
201 Matoka, K. 274 Mosher, D.L. 221,
Luria, Z 137 McAdoo, H.P. 54, 189, 207 Moskowitz, G.B. 270
Lutz, S.E. 31, 46, 58, 61 McDevitts, J. 232 Mounts, N. 136
Lyde, M. 205 McDougall, P. 300 Mull, E. 52, 63
Lykken, D.T. 295 McFaul, T. 208 Mullally, S. 88
Lyons, A. 293 McGarty, C. 4, 12, 14–15, Mullen, B. 225, 270, 273, 298
Lyons, E. 19, 62, 78, 87, 90, 86–87, 91, 164, 177, 190 Mullin, P.N. 192
332 Author index
Mummendey, A. 162 Otto, S. 50 Quattrone, G.A. 105
Murdoch, N. 86 Ou, Y. 54 Quillian, L. 232
Murphy, J. 195 Ouwerkerk, J. 164, 195 Quintana, S.M. 40–1,184
Murrell, A.J. 267, 268, 274, Ovesey, L. 191
278 Owen, C. 208 Rabbie, J.M. 7, 114
Muth, J.L. 256 Oyserman, D. 195, 208 Radke, M.J. 221, 226, 232
Myers, B. 35 Radke-Yarrow, M. 221, 222
Myers, L.C. 43 Pachter, L.49, 55 Raglioni, S.S. 135
Padilla, A.M. 35, 191, 220, Ramsey, P.G. 36, 43, 45, 191
Nadleman, L. 109 223, 224 Read, S.J. 4
Nash, S. 109 Paez, D. 60 Reavis, R. 149
Neeman, J. 299 Page, M.S. 205 Rees, E.T. 85
Neiderhiser, J.M. 295 Palmer 110 Regan, M. 86, 114, 164, 177
Neisser, U. 142 Parish 112, 118, 122 Reicher, S.D. 3, 11, 13–14, 16,
Nemceff, W.P. 50 Park, B. 111, 270 60, 79, 113, 163, 190, 220,
Nemeth, C. 160, 224, 248, Parke, R.D. 42 225, 253, 269, 315
249, 250 Parsons, J.E. 34 Reid, A. 3, 85, 296
Nesdale, A.R. 236, 237, 239, Patterson, C.J. 109 Reingold, H.L. 137
236, 237 Pederson, A. 253 Reis, H.T. 110
Nesdale, D. 49, 135, 162, 182, Pelham, B.W. 205–206 Reiss, D. 295
219, 220, 221, 222, 223, Perdue, C.W. 51, 271, 294 Renninger, C.A. 227
224, 225, 228, 229, 234, Perez, S.M. 108, 118, 119 Renshaw, P.D. 300
235, 236, 237, 248, 251, Perreault, S. 162, Resnick, L.B. 207
254, 300, 303–304, 320, Perry, D.G. 55, 118, 122, 142 Reynolds, K.J. 9–10
314, 316–320, 323–324 Peterson, A.C. 318 Rhee, E. 2, 17, 29, 36, 38, 44,
Nesdale, L.S. 20, 105, 192, Pettigrew, T.F. 270, 271, 272, 52, 56, 59, 63, 91–92, 200,
194 273, 305 291
Neuberg, S.L. 85, 272 Pheterson, G.I. 123 Rholes, W.S. 33
Newcomb, T.M. 114 Phinney, J. 32, 42, 46, 53, 159, Ricci, T. 199
Newcombe, N. 232 189, 190, 191, 195, 197, Rice, A.S. 223
Newman, L.S. 5, 33, 45, 52, 199, 207, 209 Rich, M.C. 229
59 Piaget, J. 146, 160, 161, 165 Riek, B.M. 315, 320, 322–324
Newman, M.A. 35, 39 Pichevin, M.F. 119, 149 Riley, S. 110
Nicole, S. 292 Pine, C.J. 39 Rim, Y. 224, 249, 250
Nielson Media Research Pinter, B. 268 Ring, K. 6
138 Plant, E.A. 305 Ríos, D.I. 205
Nier, J. 276 Plomin, R. 295 Robalo, E.M. 237
Nijsten, C. 198 Pomare, M. 274, 278 Roberson, J.K. 227
Noel, J.G. 51, 253 Pomerantz, E.M. 5, 55, 87 Robertson, L.S. 86
Norcliffe, H. 55 Pomerleau, A. 137 Rocha, S.A. 237
Norton, A.J. 268 Porter, L.E. 196 Rochat, P. 3
Nurious, P. 57 Porter, J.D.R. 221, 222 Rodriquez, S. 45, 47, 59
Nuttin, J.R. 114 Porter, J.R. 52, 110 Roman, R.J. 200
Postmes, T. 253 Romero, V. 117, 295
Oaker, G. 51 Pouliasi, K. 206 Rompf, W.J. 223
Oakes, P.J. 11–16, 79, 86–87, Poulin-Dubois, D. 109–110, Romney, A.K. 47
91, 93, 113–114, 144, 163, 113, 118, 122, 124, 298 Roosens, E. 194, 199
164, 177, 190, 193, 220, Powlishta, K. 18–19, 58, 90, Rose, H. 108
225, 228, 269, 315 106, 108–122, 124, 136, Rosenberg, M. 192, 197
O’Brien, M. 137 141, 147, 149, 162, 192, Rosenthal, D.A. 199
Ocampo, J.A. 37 248, 249, 254, 256, 259, Ross, L. 4
Ogbu, J. 59 293, 295, 298–299, 317, Rosch, E. 12
Ogilvie, D.M. 46 320–324 Rosenfield, D. 220, 221, 224
Ohana, J. 192 Pratto, F. 255 Rosewarne, D.L. 248
Oljenik, A.B. 138 Prawat, R.S. 122 Rotenberg, K.J. 200, 205
Onorato, R.S. 17, 19, 90, Prentice, D.A. 135 Rothbart, M. 111, 270
113–115, 135, 141, 143, Proshansky, H.M. 220, 224, Rotheram, M. 159, 189, 190,
147, 183, 209, 269 232 197
Oppenheim, D.B. 226 Prout, A. 300 Rowley, S.A.J. 42, 45–46,
Otten, S. 270, 294 Provenanzo, F.J. 137 49–51, 53, 194, 195
 Author index 333
Ruderman, A.J. 43 Sechrist, G.B. 293, 297, 304 Sprafkin, C. 44, 86, 107, 112,
Rubin, J.Z. 137 Sedikides, C. 270, 295, 297 118, 121, 136–137
Rubin, L.R. 318 Seidman, E. 34, Stangor, C. 29, 37, 43, 44, 56,
Rubin, M. 31, 54, 112, 228, Sellers, R.M. 42, 45–46, 49, 85, 111, 140, 195, 293, 294,
295 50, 53, 194, 195 297, 304
Ruble, D.N. 2, 17, 29, 31, Selman, R.L. 231 Starer, R. 114, 123
33–40, 43–46, 49–52, Semaj, L.T. 37, 40, 46, 52, Steele, C.M. 3, 53, 54, 57, 59,
55–56, 58–59, 61–63, 223, 230 195
87, 91–92, 109–111, 118, Semin, G.R. 270 Steers, W.N. 205
140, 159, 197–198, 220, Sen, M.G. 107, 109–110, 113, Steinberg, L. 136
248, 291, 292, 293, 294, 118, 122, 124, 298 Stennes, L. 110
314–318, 320–322 Serbin, L.A. 44, 58, 86, Stephan, W.G. 220, 221, 224,
Ruble, T.L. 55 106–107, 109–110, 232
Ruiz, R.A. 35, 191, 220, 223, 112–113, 118–122, 124, Stephenson, G.M. 114
224 126, 136–137, 192, 248, Sternglanz, S.H. 126
Runciman, W.G. 295 249, 254, 256, 259, 298 Stevenson, H.W. 135, 227
Ruscher, J.B. 270 Settles, B.H. 268, 276 Stevenson, N.G. 135, 227
Rust, M.C. 267, 269, 272, 275 Shaller, M. 85 Stipek, D. 36
Rutland, A. 3, 20–21, 62, 106, Shelton, J.N. 42, 45–46, 49, Stoddart, T. 318
192, 248, 249, 250, 253, 50–51, 194, 195 Stringer, P. 104
255, 258, 298, 303, 304–305, Shepherd, P.A. 107 Stroebe, W. 9
314, 316–317, 322–323 Sherif, M. 7 Strough, J. 294, 300, 301
Ryan, M.K. 18, 20, 91, 118, Sherman, F. 35, 39, 52 Sugawara, A.I. 109
120, 143, 292, 299–301, Shirley, L. 299 Suitor, J.J. 149
318–324 Short, J. 86, 161, 177 Swann, W.B. 4
Ryan, T.T. 137 Signorella, M.L. 43, 109, 118, Swim, J.K. 195
137 Szalacha, L.A. 54
Sadker, D. 148 Signorielli, N. 138 Szkrybalo, J. 36–37, 56, 58
Sadker, M. 148 Silvern, L.E. 112–113, 118
Sahm, W.B. 109 Silvia, P.J. 114, 116 Takahashi, Y. 205
Sakuma, M. 205–206 Silverman, L. 56, 58, Tajfel, H. 1–2, 6–10, 30–31,
Salvi, D. 270 Simon, B. 164, 180 42, 51–52, 56, 60, 79, 104,
Sanford, R.N. 220 Sinclair, A. 105, 114 117, 144, 147–148, 160,
Sani, F. 17, 18, 41, 43, 44, 62, Singelis, T.M. 195 161, 162, 165, 225, 190,
78–80, 86, 88, 108, 115, Singer, L.T. 107 193, 194, 220, 224, 228,
121, 160, 165, 175, 177, Skerry, S.A. 29, 38, 47 233, 234, 236, 248, 249,
192, 248, 251, 293, 295–297, Skinner, M. 114 250, 251, 267, 269, 270,
302, 306, 314–317, Slaby, R.G. 56 272, 295, 321, 322
319–321, 324–325 Smetana, J.G. 56 Tavris, C. 139
Santos, L.J. 195 Smith., C. 225, 270, 298 Taylor, A. 295
Sato, T. 199 Smith, E.R. 195, 270 Taylor, D.M. 196
Sauer, L.E. 277 Smith, P.B. 195 Taylor, M.G. 118
Schaefer, D.F. 232 Smith, M.A. 42, 45–46, Taylor, S. 43, 111, 113
Scheier, M.F. 3, 258 49–51, 53, 107, 194, 195 Teasley, S.D. 207
Schlenker, B. 4 Snyder, M. 4, 109 Teeäär, A. 199
Schlopler, J. 268 Sockloff, A. 35, 107, 109 Tein, J.Y. 137
Schmidt, C.R. 136 Sohn, M. 224, 228 Tellegen, G. 200
Schmitt, K.L. 56 Sorrell, G.T. 299, 301 Teplin, L.A. 223, 226
Schmitt, M.T. 54, 114, 116, Spanier, G.G. 268 Tesser, A. 4
194 Sparks, K. 298 Thijs, J. 195–197, 202, 204
Schnake, S.B. 270 Spears, R. 92, 114, 164, 180, Thoits, P.A. 30–31,
Schofield, J.W. 136 232, 251, 253, 269 Thompson, E.P. 294
Scholte, R.H.J. 300 Spears-Brown, C. 62 Thompson, L. 207
Schroeder, C. 229 Spence, J.T. 55 Thomson, S. 86, 92–93, 96,
Schwartz, J.L.K. 255, 257, Spencer, M.B. 40, 49, 52, 253
258 301 Thompson, S.K. 35, 107, 109,
Schwarzwald, J. 232 Spencer, S. 51, 54, 59 136
Scodel, A. 221 Spender, D. 149 Thompson, R.H.T. 223
Sears, D.O. 5 Spiaggia, M. 232 Thorne, B. 103, 114, 136, 137,
Seavey, C. 114, 228 Spielman, D.A. 198, 248 141
334 Author index
Todor, N.L. 114, 123 198, 223, 224, 225, 227, Wicklund, R.A. 259
Tonick, I.J. 126 234, 248 Wigfield, A. 57
Trafimov, D. 206, 305 Vener, A.M. 109 Wilder, D. 30, 51, 105, 114,
Trager, H.G. 221, 222, 226 Verkuyten, M. 19–20, 39, 270, 272
Trautner, H.M. 4, 6, 109, 44, 135, 162, 195–204, Wildschut, T. 268
159 206, 208, 248, 250, 293, Wilke, H. 164
Triandis, H.C. 195, 199, 206, 301–302, 315, 320–321, 324 Wilkens, G. 7, 114
305 Vevea, J.L. 295 Wilkes, A.L. 9, 144
Trolier, T.K. 85 Villareal, M.J. 195, 199 Willer, S.C. 47
Tropp, L.R. 271 Virshup, L.K. 30–31 Williams, J.A. 193, 234, 248
Tucker, G.R. 229 Vitaro, F. 299 Williams, J.E. 86, 104, 110,
Tur-Kaspa, M. 232 Voils, C.I. 296 122, 137, 223, 227, 232
Turiel, E. 318 Vollebergh, W.A. 200 Willis, H. 31
Turner, J.C. 1–2, 7, 9–17, 19, Von Hippel, W. 270 Wills, T.A. 3
31, 42, 51–52, 60, 79, Voydanoff, P. 277 Wilson, H. 87, 160, 176
86–87, 90–91, 93, 104, Wilson, M.N. 39
113–115, 117, 123, 135, Walker, I. 253 Wong, D.L. 148
139, 141, 143–144, Walker-Andrews, A.S. 135 Wood, C.H. 110
147–148, 162, 163, 164, Wann, D.L. 51, 162, 232, 253 Wootton-Millward, L. 164
177, 183, 190, 193, 194, Ward, C. 276 Wothke, W. 280
209, 220, 225, 228, 236, Wasik, B.H 189, 207 Wright, S. 110, 196
251, 269, 270, 272, 273, Weber, M. 194 Wurf, E. 142
302, 315 Webster, D.M. 294 Wyman, H. 58
Tyler, R.B. 51, 271, 294 Weil, A.M. 160, 161, 165
Tyler, T.R. 91 Weiland, A. 223 Yarkin-Levin, K. 4
Tyrrell, D.J. 119 Weinraub, M. 35, 107, 109 Ybarra, O. 232
Weinreich, P. 191 Yee, M.D. 43, 44, 50–51, 105,
Uleman, J.S. 200 Welch-Ross, M.K. 136 107, 112, 118–119, 121,
Wellman, H.M. 33, 78 162, 192, 229, 234, 239,
Vaillancourt, T. 300 Wentura, D. 294 248, 249, 251
Validzic, A. 274 Wetherell, M.S. 11, 13–14, Yeeles, C. 254, 258
Van Ausdale, D.V. 49 16, 60, 79, 104, 113, 163, Yip, T. 53
Van de Geer, J.P. 160, 161, 190, 220, 225, 228, 269, Young, H. 253
165 315 Yuill, N. 92–93, 96, 254
Van den Heuvel, H. 200 White, D.C. 112, 118, Yum, N. 205
Van de Wielen, C. 196 121–122 Yzerbyt, V.Y. 237
Van Knippenberg, A.F.M. White, D.R. 192, 248, 249,
118, 164 254, 256, 259 Zalk, S.R. 112, 118, 122, 224,
Van Lieshout, C.F.M. 300 White, G.D. 272, 277 228
Vartanian, L.R. 112, 117, 122 White, L.K. 268 Zecharia, D. 297
Vaughan, G. 36, 104, 191, Wichstrom, L. 318 Zinser, O. 229
Subject index

Aboud’s socio-cognitive theory 222, 223, 224, Conservation 37, 38, 223, 231
247, 248, 249, 250, 251, 252, 260 Constancy 35, 37, 38, 39, 40, 44, 48, 52, 56, 57,
Accentuation effect 15 58, 96, 139, 140, 151, 223, 230, 293
Accentuation of intergroup differences 18, 86, Contact hypothesis 271, 272, 273, 275, 279
89, 103, 105, 106, 108–111, 125, 144, 147, Contextual variability of stereotypes 77,
150, 162, 235, 239 85–91, 296
Accentuation of self-stereotypical Cooperativeness 277, 278
characteristics 92, 93–94 Cross-ethnic identification 53, 194, 226
Accentuation of withingroup similarities 103, Cultural characteristics 190, 195
105, 106, 108–111, 125, 144, 147,162, Cultural knowledge 198
203, 206, 269
Accessibility 13, 45, 115, 144, 146, 151, 207, Decategorisation 272, 273, 274, 322
294 Depersonalisation 15, 16, 163, 178, 203, 304,
Affective distinctiveness 170 305
Allocentrism 199, 200 Depression 55, 318
Anti-individualistic social psychology 6, 192 Discrimination 19, 48, 54, 253
Attraction 16, 123, 125, 269, 273, 317 Domain-related expertise 85, 296
Authoritarian personality 220 Dual identity 273
Automatic processing 108
Ecological theory 192
Black sheep effect 60, 237, 239
Ego identity formation 31
Body image bias 255–256, 257, 258
Essentialism 41
Eriksonian theory 31, 191
Categorisation – see Social categorisation. Ethnic awareness 35–36, 226, 227, 293, 316
Child as social tactician 21, 261 Ethnic identity 19, 31–32, 34, 36–38, 40, 42,
Children’s sense of “we” – see Identification 45, 46, 47, 48, 56, 63, 84, 189–210, 229,
with social groups 253, 304, 321
Co-construction of identities 97, 297 Ethnic norms 59
Cognitive development 3, 6, 20, 21, 29, 30, 37, Ethic preference 53, 226, 228
38, 79, 84, 85, 90, 96, 124, 125, 146, 147, Ethnocentrism 7, 8, 16
150, 151, 191, 207, 209, 222, 223, 230, Ethnographic research 34, 97
231, 233, 236, 248, 251, 291, 292, 293, Expertise – see Domain-related expertise
296, 303, 306, 316, 317, 318, 319
Collective identity – see Social identity
Collectivism 195, 199, 200, 205, 206, 208, 209, False self 142
305 Families 34, 54, 62, 96, 192, 199, 267–285, 297,
Common ingroup identity model 21, 267, 269, 303, 305, 313, 314, 315, 322, 324
272, 276, 277, 278, 279, 281, 283, 284, Fit 13, 151, 163, 193
305 comparative 14, 96, 144, 146, 149, 318
Comparative context 77, 85–91, 95, 141, 163, normative 14, 144, 146, 149, 318
164, 176, 190, 209, 294, 300, 302 Forced-choice methods 233
Competence/performance distinction 303
Confounding of ingroup favouritism and Gender differences 55, 58, 89, 110, 111, 117,
outgroup prejudice 61, 113, 220, 303 120, 125, 135–152, 248–249, 254, 322
336 Subject index
Gender identity 18,34, 36–38, 42, 44, 45, 46, Intergroup differentiation 52
56, 57, 58, 63, 77, 84, 87–90, 135–152, Intergroup discrimination 8, 9
209, 300, 301, 321 Interaction with peers 34, 96, 196, 204, 292,
Gender norms 59, 61, 137, 301 297
Gender role development 106, 107, 209 Internalisation of social identity – see
theories of 139–140 Identification with social groups
Gender role knowledge 58, 109, 119, 121, 123, Intragroup comparisons 176
145
Gender schematic processing theory 84, 116, Kohlberg’s theory 139–140, 147
140, 141
Gender segregation 112, 124, 125, 135, 136, Legitimate/illegitimate differential status
147, 298, 320 10–11, 148–9, 194
Genetic social psychology 192
Group distinctiveness 269
Mediation analysis 278–279
Metacontrast principle 14, 96, 144, 163,
Identification with social groups 17, 18, 29, 32,
304
33–37, 53, 77, 91–95, 139, 146, 147, 161,
Minimal group paradigm 8, 10, 11, 31, 104,
162, 164, 168, 175, 176, 178, 182, 183,
105, 106, 120, 124, 147, 225, 233, 234,
184, 191, 192, 193, 195, 198, 200, 201,
235, 236, 248, 315
208, 233, 250, 275, 292, 297, 302, 315,
Moral reasoning 231, 232
316, 320
Multicultural education 195
Identity-related beliefs 78–85, 316
Multilevel modeling 201, 202
Individual mobility 10
Mutual intergroup differentiation 274
Immigrant children 45, 46, 47, 48, 50, 52, 54,
59, 189–210, 253, 315, 320
Implicit association test 257–258 Name-calling 196, 197, 224, 320
Implicit measures 206, 254–258 National identity 19, 78, 79, 90, 106, 149,
Impression management – see Self- 159–184, 302, 316, 324
presentation Negotiating identities 34, 96
Individualism 195, 205 Normative features of group members, see
Infants 35, 36, 107, 109, 135, 136, 137, also Social identity, children’s
151 conceptions of 78, 79
Information-seeking 56–57, 228 Norms 16, 221, 247, 252, 253, 254, 255, 258,
Ingroup bias 7–10, 18, 19, 20, 50, 51–52, 55, 259, 275, 298, 299, 302, 304
60, 61, 103, 104, 105, 106, 112–113, 114,
115, 117, 118, 119, 120, 122, 123, 124, Outgroup derogation – see Prejudice
125, 147, 148 161, 163, 164, 175, 197, Outgroup favouritism 10, 249, 253
219, 223, 228, 247, 249, 250, 251, 253, Outgroup homogeneity 103, 105, 106, 111,
259, 267, 272, 274, 275, 276, 294, 295, 116, 118, 125
303, 314, 322, 323
Ingroup denigration 168 Peer group 96, 292, 297
Ingroup deviants 60, 237, 239 Perceiver readiness 13
Ingroup favouritism 31, 51, 55, 61, 103, 104, Permeable/impermeable group boundaries 10,
105, 106, 112–113, 114, 115, 117, 118, 96, 105, 148, 149, 151, 230
119, 120, 122, 123, 124, 125, 147, 148, Personal/group discrimination discrepancy
160, 161, 162, 163, 165, 166, 168, 175, 196, 197
192, 194, 198, 220, 222, 224, 225, 228, Personal identity 1, 2, 12, 18, 63, 96, 113, 142,
229, 234, 239, 248, 269, 270, 294, 295, 143, 150, 193, 204, 206, 228, 269, 295,
297, 300, 302 297, 320
Ingroup homogeneity 163, 164, 176, 177, Piagetian theory 3, 38, 146, 147
178–180, 181–182 Positive distinctiveness 8, 148, 149,151,
Ingroup identification – see Identification with 162, 163, 168, 170, 172, 177, 197, 226,
social groups 304
Ingroup norms 60 Prejudice 1, 20, 54, 60, 61–63, 122, 136, 141,
Intergroup comparisons 79, 87 161, 162, 190, 191, 194, 195–198, 200,
Intergroup competition 7 204, 207, 208, 209, 219–240, 247, 248,
Intergroup conflict 7, 60, 135, 232, 236, 268, 249, 251, 252, 253, 255, 260, 261, 269,
272, 279, 281, 300 303, 304, 305, 316, 317, 320, 322, 323,
children’s explanations for 79–81 324
Intergroup contact – see Contact hypothesis insider’s perspective 195
 Subject index 337
Prototype 14 Social categories as constitutive of self 1, 33
Prototypicality 14, 16, 203, 270 Social categorisation 6–17, 61, 62, 104, 105,
Psychoanalytic theory 139 107, 192, 227, 267, 271, 272, 273
Public self-focus 258–260 Social change beliefs 10
Social comparison 1, 9, 31, 52, 54, 59, 61, 62,
Racial identity – see Ethnic identity 79, 87, 162, 209, 226, 234, 292, 295
Racism, see also Prejudice 62, 189, 194 Social competition 11, 194, 234
Recategorisation 272, 273, 274, 275, 284, 322 Social context 1,5, 6, 7, 15, 39, 44, 46, 62, 63,
Referent informational influence 16 103, 113–115, 125, 126, 144, 147, 148,
Relational self 142 149, 150, 161, 189–210, 224, 235, 249,
Remembered self 142 251, 253, 274, 291, 293, 294, 296, 297,
299, 301, 303, 306, 314, 315, 317, 318,
Scaffolding of identities 84 321, 323
Schools 189, 192, 195, 201, 202, 294, 296, 297, Social creativity 10, 194, 297
298 Social exclusion 196, 259, 260
Self-affirmation 53 Social identity
Self as process 18–19, 135, 141–145, 150, 209 across the lifespan 4, 22, 323–324
Self as structure 140, 141, 201, 206, 209, 302 as a multidimensional construct 17, 30, 42,
Self-categorisation 9, 94, 190, 193, 195, 62, 63, 209
202–206, 208, 235, 236, 239, 291, 294, and engagement in school 53, 59
302, 303, 304, 306, 316, 323, 324 children’s conceptions of 17, 18, 30, 33, 35,
Self-categorisation theory 18, 19, 20, 60, 78, 37–38, 40–41, 78–85, 316
86, 88, 96, 113, 115, 125, 141–146, 159, consequences of developmental change in
161, 163–164, 176, 178, 182–185, 190, 51– 63
192, 202, 204, 205, 207, 209, 220, 225, motivational consequences and 55–63, 147
228, 252, 261, 269, 273, 292, 297, 302, Social identity development theory 20, 220,
303, 304, 306, 315, 318, 321, 325 225, 226–240, 303,
Self-concept 1–4, 9, 12, 17, 20, 29, 30, 33, 34, Social identity theory 1, 2, 6–17, 19, 20, 31,
42, 45, 46, 51, 63, 87, 142–145, 192, 197, 42, 51, 55, 56, 58, 60, 78, 104, 105, 113,
199, 205, 209, 291, 292, 293, 294, 295, 117, 124–125, 159, 162–163, 175,176,
296, 297, 300 182–185, 190, 192, 197, 198, 204,
Self-descriptions 33, 91, 200, 202, 205, 254, 207, 208, 209, 220, 225–226, 228, 247,
297, 302 251–252, 261, 269, 295, 296, 302, 303,
Self-efficacy 53, 55 304, 306, 315, 316, 325
Self-esteem 2, 9, 17, 29, 51, 52–55, 57, 104, Social learning theory 46, 139, 140, 141, 142,
105, 110, 117, 125, 149, 162, 175, 192, 145, 146, 147, 151
195, 225, 226, 228, 234, 251, 294, 315, Social mobility beliefs 10
320, 322 Social representations 14
collective 49, 148, 151, 196, 197, 198, 199, Social status 22, 39, 40, 48, 49, 53, 62, 105,
200, 201–2, 203, 204, 208, 295 118, 162, 182, 183, 190, 192, 194, 198,
personal 49, 50, 204, 295 207, 208, 226, 232, 234, 235, 236, 239,
Self-presentation 45, 253, 254, 258, 260, 304, 251, 268, 271, 275, 296, 298, 322
306, 318, 323 Socialisation 46, 49, 54, 62, 84, 89, 125, 135,
Self-recognition 299 140, 152, 190, 198, 199, 221, 255
Self-regulatory processes 46, 252–261, 304, Stepfamily conflict 267–285, 324
306, 317 Stereotypes 49, 56, 58, 59, 63, 85–91, 105, 109,
Self-representations 2–3, 4 110, 111, 114, 121, 123, 136, 137–138,
Self-schemas 2, 140 140, 143, 145, 151, 152, 161, 164, 194,
Sex differences – see Gender differences 197, 198, 228, 229, 237, 270, 271, 297,
Social categories 29, 30 298, 300, 301, 317, 320, 325
awareness of 35–36, 319 cognitive structural conception 77, 85, 86,
centrality of in self-conception 34, 45–49, 90
193, 203, 204, 205, 206, 294 contextual variability in 71, 85–91, 296
evaluation of 49– 51 SCT conception 86, 90–91
identification with – see Identification with Stereotyping 43, 60,62, 103, 122, 162, 219,
social groups 302, 321
knowledge of 37–38, 48–49, 295 ingroup 177–178
salience of in self definition 43– 45, 103, self 93–94, 114, 152, 178, 302
144, 149, 176, 204, 229, 250, 269, 273, Stigmatization 53, 191,194
294 Stroop effect 256
338 Subject index
Structural equation analysis 277–283 Uncertainty reduction 31, 294, 306
Subjective identification with social groups –
see Identification with social groups Victimization 194, 204, 205, 219, 229
Symbolic interactionism 190 Vygotskian theory 97, 297

Theory of mind 84, 94 Working self 142