Self-Regulatory Mechanisms Governing Gender Development

Author(s): Kay Bussey and Albert Bandura

Source: Child Development, Vol. 63, No. 5 (Oct., 1992), pp. 1236-1250
Published by: Wiley on behalf of the Society for Research in Child Development
Stable URL: https://www.jstor.org/stable/1131530
Accessed: 28-08-2018 10:00 UTC

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/1131530?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Society for Research in Child Development, Wiley are collaborating with JSTOR to
digitize, preserve and extend access to Child Development

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
Self-Regulatory Mechanisms Governing
Gender Development

Kay Bussey
Macquarie University

Albert Bandura

Stanford University

BUSSEY, KAY, and BANDURA, ALBERT. Self-Regulatory Mechanisms Governing Gender

ment. CHILD DEVELOPMENT, 1992, 63,1236-1250. This study tested predictions about develo
of gender-related thought and action from social cognitive theory. Children at 4 levels of
constancy were assessed for their gender knowledge, personal gender standards, and g
linked behavior under different situational conditions. Irrespective of gender constanc
all children engaged in more same-sex than cross-sex typed behavior. Younger children
in a gender stereotypic manner to peers' gender-linked behavior but did not regula
own behavior on the basis of personal gender standards. Older children exhibited su
self-regulatory guidance based on personal standards. They expressed anticipatory self-
for same-sex typed behavior and self-criticism for cross-sex typed behavior. Their anti
self-sanctions, in turn, predicted their actual gender-linked behavior. Neither gender kn
nor gender constancy predicted gender-linked behavior. These results lend support
cognitive theory that evaluation and regulation of gender-linked conduct shifts develop
from anticipatory social sanctions to anticipatory self-sanctions rooted in personal stand

1983). For example, children prefer same-

Because so much of human experience
sex
is affected by gender differentiation, the toys (Carter & Levy, 1988; Marcus &
pro-
cesses governing gender developmentOverton,
con- 1978), imitate same-sex models
tinue to be the subject of much develop-
(Bussey & Bandura, 1984), and reward peers
mental theorizing and research. Proponents
for gender-appropriate behavior before they
of cognitive-developmental theory (Kohl-
have fully attained gender constancy (Lamb
berg, 1966) advanced gender constancy as
& Roopnarine, 1979). Moreover, growing
the driving force guiding young children's
awareness of gender constancy does not in-
crease children's preferences for same-sex
gender-related behavior (Stangor & Ruble,
1987). According to this theory, onceroles
chil-and activities (Marcus & Overton,
dren achieve gender constancy-a concep-
1978; Smetana & Letourneau, 1984). Thus,
tion of their own gender as fixed andfactors
irre- other than gender constancy seem to
versible-they positively value and seek tochildren's gender-linked behavior.
guide
adopt only those behaviors congruent with
the gender concept they have acquired. Because of its limitations, cognitive-
Children are not expected to adopt sex-
developmental theory has been modified
typed behaviors consistently until afterand
they
extended within the explanatory frame-
have labeled themselves unalterably work
as a of gender schema theory (Carter &
boy or a girl, which usually is not achieved
Levy, 1988; Martin & Halverson, 1981). This
until about age 6. However, many studies
theory bears many similarities to cognitive-
have failed to corroborate the link between developmental theory, but departs from it in
children's attainment of gender constancy
two significant ways. First, the attainment of
and their gender-linked conduct (Huston,complete gender constancy is not consid-

This research was supported by Public Health Research grant MH-5162-25 from the Na-
tional Institute of Mental Health to Albert Bandura, and Australian Research Council grant
A78930570 to Kay Bussey. We are grateful to Kathleen Davies, Elizabeth Grimbeek, and Gwyn
Wachtel for assisting with the data collection, and to the children at Bing Nursery School for
their participation in the study. We thank Barry Zimmerman for helpful comments on an earlier
draft of the manuscript. Requests for reprints should be sent either to Kay Bussey, School of
Behavioral Sciences, Macquarie University, North Ryde, Sydney, Australia, 2109, or to Albert
Bandura, Department of Psychology, Stanford University, Stanford, CA 94305-2130.

[Child Development, 1992, 63, 1236-1250. ? 1992 by the Society for Research in Child Development, Inc.
All rights reserved. 0009-3920/92/6305-0009$01.00]

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 Bussey and Bandura 1237

der-related information. However, the rela-

ered necessary to motivate and guide chil-
dren's gender-linked behavior (Martin tion
& of gender schematization to children's
Halverson, 1981; Stangor & Ruble, 1987).
gender preferences has been problematic.
The findings have been inconsistent across
Second, the information-processing func-
tions of the schema are given greater em-
different measures of gender schematization
phasis in gender schema theory than in and across age groups (Carter & Levy, 1988;
Kohlberg's theory (Bem, 1981; Martin Edwards
& & Spence, 1987; Signorella, 1987).
Halverson, 1981). Moreover, Signorella (1989) has noted that
children's knowledge about gender-related
Gender schema theory initially pro-
stereotypes is unlikely to explain gender de-
posed that only the attainment of gender
velopment, because most young children
identity was necessary for the acquisition
"know" the gender stereotypes but differ in
of gender-linked behavior (Martin & Hal-
gender-linked conduct. Apparently, gender
verson, 1981). Once children could label
knowledge is not the main determinant of
their own gender and that of others, they
children's gender-linked conduct. Both cog-
were expected to behave in ways consistent
nitive-developmental theory and gender
with traditional gender roles. In keeping
schema theory have focused on gender con-
with these expectations, Fagot and Leinbach
ceptions, but neither devotes much attention
(1989) found that children who acquired
to the translation of gender-linked concep-
gender-labeling skills prior to 28 months
tions to gender-linked conduct. Nor do they
(early labelers) were more likely to play with
specify the motivating mechanism for acting
traditional gender-linked toys than those
in accordance with a conception (Bandura,
who had not yet mastered gender labeling
1986). Knowing a stereotype does not ne-
by that age (late labelers). However, it re-
cessarily mean that one strives to behave
mains unclear whether gender labeling and
in accordance with it. For example, self-
gender-linked preferences are coeffects of
conception as elderly does not enhance val-
social influences or are causally linked.
uation and eager adoption of the negative
Early labelers may simply be more preco-
stereotypic behavior of old age.
cious than late labelers, demonstrating gen-
der-related play preferences learned earlier
Social cognitive theory provides a third
from parents, peers, and the media. More-
perspective on gender development and the
over, at this age, children are unable to label
mechanisms governing the motivation and
accurately the gender linkage of toys (Wein-
regulation of gender-linked behavior (Ban-
raub et al., 1984), regardless of their ability
dura, 1986, 1989). This theory specifies the
to label accurately the gender of persons.
multifaceted determinants of thought and
Hence, knowledge of gender-linked label-
action and the regulative mechanisms by
ing cannot explain the early phases of gen-
which they are linked. Social cognitive the-
der development. To complicate matters
ory does not require that action be depen-
even further, Martin and Little (1990) have
dent on gender knowledge in the early
recently found that the strongest correlate of
phase of gender development. Because gen-
gender-linked preference was gender stabil-
der-related cues are available for gender la-
ity. A weaker relation was found with gender
beling, in this view, children learn to label
identity, but none was obtained between
their own and others' gender before they
gender consistency and gender-linked pref-
learn to label and categorize objects, activi-
erence. Although gender schema theory dif-
ties, tasks, and roles that, of themselves,
fers from Kohlberg's theory in the level of
have no inherent gender linkage. It is from
gender understanding that is considered
children's social and observational experi-
necessary for the acquisition of gender-
linked behavior, the measures of gender ences that gender-linked knowledge
conception remain the same. emerges. As children develop stronger gen-
der-linked preferences, their knowledge of
With regard to the information- the constellations of attributes that are
processing aspect of gender schema theory,
linked to gender increases. In social cogni-
the more salient or available the schema,tive
thetheory (Bandura, 1986), children's grow-
more individuals are expected to attend to,
ing cognitive competence is but one factor
encode, represent, and retrieve information
involved in their gender-related develop-
relevant to gender. Research conducted ment. Proximal social influences of parents,
within this conceptual framework has added
teachers, and peers, as well as distal social
to our understanding of how gender- and symbolic influences from the mass me-
schematic processing affects allocationdia
ofand cultural institutions, all serve to pro-
attention, organization, and memory of mote
gen- gender development. In this theory of

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 1238 Child Development

triadic reciprocal causation, the social envi-

both a standard-matching function and an af-
ronment, children's knowledge structures fective self-reactive function. Research con-
ducted in different domains reveals that
and cognitive capabilities, and their behav-
ior interact to produce gender-related stan-
both functions are necessary in the motiva-
dards and action. tion and regulation of conduct (Bandura,
1991b).
Viewed from the sociocognitive theoret-
ical perspective, it is not surprising that chil-
The self-regulatory mechanisms speci-
dren exhibit gender-linked preferences fied in social cognitive theory have been
prior to achieving gender constancy. Fromshown to operate as important motivators
the moment of birth children are socialized and regulators of children's conduct in other
according to their gender (Rheingold & major domains of functioning. For example,
Cook, 1975). Parents explicitly and implic-children exhibit self-reactive control of
itly convey to their children gender- transgressive conduct (Bandura, 1991a; Gru-
appropriate behavior. As a consequence, sec & Kuczynski, 1977; Perry, Perry, Bussey,
children act in accord with gender-linked English, & Arnold, 1980), aggressive pat-
stereotypes before they are fully cognizant terns of behavior (Perry & Bussey, 1977),
of the culturally derived gender-linked ste- and of the course of their cognitive develop-
reotypes and before they have achieved gen- ment (Zimmerman, 1989). However, the reg-
der constancy (Blakemore, LaRue, & Olej-ulative role of self-influence through per-
nik, 1979; Weinraub et al., 1984). Social sonal standards in gender-related behavior
cognitive theory posits that, in the course of has not been systematically examined.
development, the regulation of behavior
shifts from predominantly external sanctions The present study was primarily de-
and direction to gradual substitution of inter- signed to test predictions from social cogni-
nal sanctions and mandates rooted in per- tive theory regarding the emergence and
sonal standards (Bandura, 1986). Initially, regulation of gendered thought and action.
behavior is self-regulated on the basis of an- However, for comparative interest, the con-
ticipatory outcomes mediated by the social tributions of factors emphasized in cog-
environment. With increasing experience, nitive-developmental theory and gender
social knowledge, and cognitive develop- schema theory, such as gender labeling, gen-
ment, children construct their own personal der constancy, and gender-linked knowl-
standards relating to gender-linked conduct. edge, were also examined. Children were
Such conduct is then motivated and regu- selected at four levels of gender conception,
lated mainly by the exercise of self-reactive ranging from gender labeling to gender con-
influence. stancy, and their gender-related standards
and conduct were assessed. Evaluative stan-
Evaluative self-reaction is the mecha-
dards are manifested not only in self-
nism whereby standards motivate and regu-
reactions but in the sanctions applied to the
late conduct anticipatorily (Bandura, 1986,
behavior of others. Therefore, children's
1991a). After an internalized self-regulative
gender-linked knowledge and social sanc-
mechanism is developed through the com-
tions toward peers' behavior that is tradition-
bined influence of modeling, tuition, eval-
ally regarded as cross-sex were also mea-
uative feedback, and environmental struc-sured.
turing, children guide their conduct by
sanctions they apply to themselves. They do Based on social cognitive theory of gen-
things that give them self-satisfaction and ader development, it was predicted that chil-
sense of self-worth. They refrain from be- dren would be aware of social sanctions for
having in ways that violate their standards
sex-typed behavior and behave in gender-
to avoid self-censure. The standards provide related ways before they displayed anticipa-
the guidance; the anticipatory self-sanctionstory self-approval for same-sex typed behav-
the motivators. Self-sanctions thus keep con-ior and self-criticism for their own cross-
duct in line with internal standards. Devel- sex typed behavior. It was further hypothe-
opmentally, children learn to evaluate andsized that after children adopted gender
regulate gender-linked conduct on the basis standards, their anticipatory self-sanctions
of external anticipatory sanctions beforewould predict their gender-linked conduct.
they do so in terms of anticipatory self-In accord with findings of previous tests of
sanctions rooted in personal standards. cognitive-developmental theory, children at
Whereas gender schema theory emphasizes higher levels of gender constancy were not
conception matching as the primary regula- expected to be more likely to engage in sa-
tive process, social cognitive theory posits me-sex behavior and shun cross-sex behav-

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 Bussey and Bandura 1239

ior than children at lower levels of gender

test. It was used in this study to establish
constancy. In gender schema theory, genderchildren's gender labeling competence that
may have been masked by the Slaby and
identity, and more recently gender stability,
rather than the complete attainment of gen-
Frey (1975) test. All children exhibited accu-
der constancy, are considered necessaryrate
to gender labeling. Consequently, four
guide children's gender-linked behavior.
progressive levels of gender conception
Hence, from this perspective, children whowere established: (1) no gender identity, but
accurate gender labeling; (2) gender iden-
had attained the lower levels of gender con-
ception, gender identity and gender stabil-
tity; (3) gender identity and gender stability;
ity, would be expected to engage in more
(4) gender identity, gender stability, and
same-sex behavior than cross-sex behavior.
complete or almost complete gender con-
sistency (one out of the three questions
Method incorrect).

Subjects Gender-Linked Self-Evaluative Standards

Subjects were 40 nursery school chil- A second experimenter, blind to the
dren (20 girls and 20 boys) from predomi- child's level of gender conception, brought
nantly middle-class families. They ranged in each child individually to the testing room.
age from 2.5 to 4.7 years, with a mean age of The experimenter explained that she was
3.5 years. Equal numbers of boys and girls helping a friend set up a toy store. Her friend
were selected at one of four levels of gender wanted to know which toys children liked to
conception. Three female experimenters play with so she could stock them in her toy
conducted different phases of the study. store. It was further explained that only the
toy store lady, not the experimenter, needed
All children participated in two sessions to know how the children would feel about
approximately 3 to 5 days apart. In the first
playing with each of the toys. The children
session, children's level of gender concep-
were told that their responses would be re-
tion was measured. In the second session,
corded automatically and anonymously on
gender-related knowledge, evaluative stan-
a "computer" and that the experimenter
dards, and conduct were measured. One ex-
would sit with her back to the computer so
perimenter conducted the first session, and a
that she was unaware of their responses. It
second experimenter conducted the second
was also explained that the toy store lady
session. A third experimenter recorded chil-
was asking many children about their reac-
dren's evaluative responses and conduct
tions to the toys and that she would not be
from behind a one-way observation mirror.
able to identify their particular responses.
Each experimenter was blind to children's
The "computer" consisted of a panel of
performances on the other aspects of the
assessment. lights and switches for reporting self-
evaluative reactions. This format enabled
Gender Conception even the youngest children in the study to
Children's level of gender conception
express their evaluative reactions privately
without
was assessed from their performances the experimenters being aware of
on the
their test,
Slaby and Frey (1975) gender constancy responses. This response procedure
which consists of three components that full control over any possible
thus provided
most children master in the following se- bias and social influence of the
experimenter
children's evaluative
quential order: (1) gender identity-- reactions.
knowledge of self and other's gender; (2)
gender stability-knowledge that gender re-were first trained to use the
Children
mains invariant across time; (3) gender lightscon-
and switches to record the nature and
sistency-knowledge that gender strength remains of their self-evaluative reactions by
invariant across situations (Fagot, 1985;
playing a ring toss game. They were taught
Slaby & Frey, 1975). Those children who to record their positive and negative evalua-
failed the gender identity component of the tive reactions by pressing a switch that acti-
test were administered the lower-level gen- vated the light behind it. When the light was
der-labeling test (Fagot, Leinbach, & Hagan, activated a buzzer sounded to inform the ex-
1986). This test requires only that the child perimenter that the child had responded.
points to the picture of a man/woman and The use of the buzzer enabled the experi-
boy/girl when the gender labels are pro- menter to sit with her back to the lights to
vided by the experimenter. Further, it does remain blind to the child's response. The ob-
not use the more complicated repeated server, also blind to the child's level of gen-
questioning procedure of the Slaby and Frey der constancy, was located behind a one-

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 1240 Child Development

way mirror and recorded the children's

were scored on a 5-point scale with 1 repre-
evaluative reactions. senting real awful and 5 representing real
great.
To measure strength of self-reactions,
Gender-Linked Behavior
five lights were mounted vertically: a large
red light, a small red light, a large orange Each child's gender-linked behavior
was measured during an unstructured play
light, a small green light, and a large green
session. The experimenter removed the box
light. The lights signified five different self-
containing the toys used for the personal
evaluative reactions using descriptors estab-
lished in pretest as those used by nursery standards measure and explained that she
school children to represent approving was and taking them to children in the room next
door. She left the other set, randomly ar-
critical affective reactions. Specifically, chil-
dren were told that they should press the ranged in one of four orders on the floor for
the child to play with while she was gone.
switch on the large green light if they antici-
pated feeling real great with themselves for
The child was told that she or he could play
having played with a particular toy. Simi-
with any of the toys while the experimenter
larly, they should activate the small green
was away and that when she returned she
would knock on the door to let the child
light if they would feel kinda great, the or-
ange light for feeling nothing special, the know that she was coming back into the
small red light for feeling kinda awful, androom. They were given the information
about the signaled entry to remove any
the large red light for feeling real awful with
themselves after having played with each possible
of external constraint on their gen-
the toys. der-linked behavior for fear that the ex-

Children were trained to respond evalu-

perimenter would walk in on them unan-
nounced and see what they were doing.
atively to their attainments with the ring toss
game until they could correctly use the The duration of play with each toy was
lights to signify approving and critical self- recorded by an observer who observed the
reactions. This preparatory training removed play sessions through a one-way mirror. The
the novelty of activating lights and ensured observer was provided with a behavior form
full understanding of their evaluative im- that listed each of the toys available for play.
port. Following instruction in the procedure, Each time the child touched a toy a timer
children's gender-linked personal standards was started and the duration of the play be-
were measured by the degree to which chil- havior was recorded. If the child played si-
dren anticipated self-satisfaction or self- multaneously with several toys, the behavior
criticism for playing with different types of durations were recorded separately for each
toys. Two parallel sets of toys (A and B), each toy. A second observer independently re-
consisting of five toys, were used. One set corded the duration of play behavior for
was used to measure personal standards and eight of the children. The interobserver
the other to measure spontaneous gender- agreement for duration of play with each of
linked behavior. Within each of the four lev- the various toys was 94%, indicating very
els of gender conception, sets A and B were high scorer reliability.
counterbalanced for assessing self-evalua-
After 3 min the experimenter returned
tive standards and gender-linked behavior
to control for any possible variation in the and explained that the children next door
needed more toys and that she would take
attributes across the two sets of toys. Toys
some to them. She removed three of the
were selected on the basis of gender-linked
toys-two same-sex and one neutral-and
ratings by adults. Each set included a dump
truck (highly masculine), robot (moderately remarked that the child could continue play-
masculine), xylophone (neutral), kitchen set ing with the remaining toys, which were
(moderately feminine), baby doll (highly both cross-sex, one highly gender-linked
feminine). The two sets of toys differed in and one moderately gender-linked toy.
color and other appearance features, but not Again, the child's spontaneous play behavior
in size. Children were shown each of the was recorded. During both of the behavior
toys separately in one of four different se- test sessions children were left alone to play
with the toys.
quences, and they registered their antici-
pated self-evaluative reaction. For half of the
Evaluative Social Reactions
children in each condition the red light was Upon her return, the second experi-
uppermost and for the other half the green menter invited the child to watch some
light was uppermost to control for any possi-
video clips with her. She explained that the
ble position affects. Evaluative reactions toy store lady had produced some videos to

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 Bussey and Bandura 1241

advertise her store on television, and thatasked to sort each of the 12 toys comprising
she would like the children's appraisal of the
the A and B sets and two highly sex-typed
videos. Four 7-year-old children, two boys toys, one feminine and one masculine, from
and two girls, enacted cross-sex preferences
the videos into one of three boxes: a boys'
following a standardized format. The videobox, a girls' box, and a boys' and girls' box.
depicted a child entering a toy store where
Pictures of either two boys, two girls, or a
six toys were displayed on a table. The toys
boy and a girl were placed in front of the
included a dump truck (highly masculine), corresponding boxes to remind the children
tool set (moderately masculine), a piano ofandthe three categories. The observer re-
puppet (both neutral), a tea set (moderately
corded the sorting choices. Items were
feminine), and a baby doll (highly feminine).
scored as correct if the toy was sorted into a
box that matched the cultural sex role stereo-
The children in the videotaped presen-
type of masculine, feminine, or neutral, as
tation examined each of the toys closely and
rated by adults prior to the study. Children
then selected a highly cross-sex typed toy.
received a score of 1 for each toy correctly
The boy selected the baby doll and played
sorted. The total possible score was 12.
with it for 21/2 min. He changed the doll's
diaper, fed her, and patted her. The girl se-
Results
lected the dump truck and played with it for
21/2 min. She put blocks in the back of the A primary aim of this study was to ex-
truck, tipped them out, and scooped them amine the influence of sanctions arising
back up again. Four videotapes were pro- from gender-linked personal standards on
duced with different girl and boy actors togender-related conduct. Hence, the first
control for any possible idiosyncrasies of analysis
the reported is of children's gender-
child actors, and to counterbalance the order
linked standards and the way in which self-
of appearances of girl and boy actors. regulatory control of their own gender-
The children watched the video clipsrelated
of conduct and that of peers changes
with age. To test predictions from social cog-
the girl and boy separately engage in cross-
sex behavior and then recorded their evalua- nitive theory concerning regulatory pro-
tive reactions using the light switches. At the cesses, a micro-level analysis of the relation
outset, a procedural check was performed to between anticipatory self-evaluative reac-
ensure that the child remembered the mean- tions and gender-linked conduct is reported.
ing of the lights and the operation of the Children's gender conceptions and their re-
switches. They all did. The response options lations to gender-linked conduct is also ex-
were the same as those used in the assess- amined.
ment of personal standards: real great, kinda
Evaluative Reactions and Age
great, nothing special, kinda awful, real To examine age differences in chil-
awful. After watching the video for approxi-
dren's evaluative self-reactions, children
mately 30 sec the child was asked, "What were divided into two equal groups on the
would this girl's/boy's friends think about
basis of their age. The younger children's
her/him playing with this doll/truck?" Themean age was 37 months, and the older chil-
child was asked to record his or her evalua-
dren's mean age was 48 months. Table 1 pre-
tions by activating the appropriate light sents the means and standard deviations for
switch. The experimenter was again seated children's affective evaluative reactions to-
with her back to the lights, blind to the re- ward gender-linked conduct as a function of
sponses. The child watched the video for a age, sex, and agent of the conduct.
further 30 sec before being asked, "What do
you think about this boy/girl playing with Self-evaluative reactions.-A 2 (age) x
the doll/truck?" Again, the child was in- 2 (sex) x 2 (gender-linked toys) analysis of
variance was performed on children's self-
structed to record his or her evaluations us-
ing the light switches. evaluative reactions. In this and all subse-
quent analyses, when significant interac-
Gender Knowledge tions were obtained, the source of the effects
The gender knowledge test was admin- were examined using the Bonferroni method
istered as the final task. The experimenter with an alpha of.05. There was a significant
explained that the toy store lady wanted to interaction for sex of subject and gender-
display the boys' toys separately from the linked toy type, F(1,36) = 32.88, p < .0001.
girls' toys in her shop. Toys played with by This interaction was, however, qualified by
both boys and girls would be displayed in a three-way interaction involving age, sex of
a third location in the store. Children were subject, and gender-linked toys, F(1,36) =

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 1242 Child Development
TABLE 1

MEANS (and Standard Deviations) FOR CHILDREN'S EVALUATIVE REACTIONS

TOWARD GENDER-LINKED CONDUCT AS A FUNCTION OF AGE,
SEX, AND THE AGENT OF THE CONDUCT

YOUNGER OLDER

EVALUATIVE REACTIONS Girls Boys Girls Boys

Self-evaluative reactions:
Feminine-linked behavior ......................... 3.9 3.6 4.3 1.9
(.7) (1.3) (1.1) (1.0)
Masculine-linked behavior ........................ 3.2 4.3 2.9 4.7
(1.2) (1.2) (1.6) (.3)
Social-evaluative reactions:
Self-reactions:
Boys' cross-sex behavior ........................ 3.0 2.7 2.8 2.5
(1.4) (1.2) (1.8) (1.5)
Girls' cross-sex behavior ......................... 3.0 3.4 2.9 2.0
(1.4) (1.3) (1.6) (1.2)
Peers' reactions:
Boys' cross-sex behavior ....................... 3.2 3.2 3.0 2.2
(1.6) (1.7) (2.1) (1.3)
Girls' cross-sex behavior ............................ 3.4 3.1 2.5 2.0
(1.3) (1.5) (1.5) (1.7)

8.07, p < .006. The interaction is gender-linked

their depicted play, and Slaby and
graphically in Figure 1. Frey (1975) found no difference between
children at Levels 1 and 2 or Levels 3 and 4
The younger children did not differ
in their sig-
attention to same-sex models. There-
nificantly in their self-evaluative
fore, to reactions
increase the size of the sample for
for gender-linked behavior. the However, as
group comparisons, children assigned to
predicted, the older children Levels
exhibited dif-
1 and 2 were combined to form Level
ferential self-evaluations for I gender-linked
and children at Levels 3 and 4 were com-
behavior. Both boys and girlsbinedexhibited an-
to form Level II. Children at Level I
ticipatory approving self-reactions for same-
had acquired gender labeling and children
sex behavior and disapproving at
self-reactions
Level II had mastered at least the gender-
for cross-sex behavior.
identity and gender-stability components of
Social evaluative reactions.-Chil- gender constancy.
dren's reactions to peers' cross-sex behavior
A 2 (gender constancy) x 2 (sex) analy-
and their judgments of how their sis peers
of variance was computed on the number
would react to such behavior were subjected
of toys correctly assigned by each subject to
to separate 2 (age) x 2 (sex of subject) x 2 girls', and both boys' and girls'
the boys',
(sex of peer) analyses of variance.category. There The only effect to attain signifi-
were no significant main or interaction ef-
cance was a main effect for level of gender
fects. The mean for the children's evaluative conception, F(1,36) = 11.26, p < .002. Chil-
reactions to peers' behavior (M = 2.79) and dren at the lowest level of gender concep-
that for assumed peers' reactions (M = 2.83)tion displayed less gender knowledge (M =
indicated that children reacted negatively to 6.00) than did children at the higher level of
both boys' and girls' cross-sex behavior andgender conception (M = 8.65). It might be
expected their peers to react in the same argued that children who placed a sex-typed
way. toy, either masculine or feminine, in the box
for both boys and girls were evidencing
Gender Knowledge flexible gender stereotypes. Hence, an anal-
Preliminary analyses revealed no sig- ysis in which sex-typed toys placed in the
nificant differences between children at neutral category were scored correct was
Levels 1 and 2 or between children at Lev- conducted. Although the mean scores were
els 3 and 4 on any of the dependent mea- raised, the pattern of results was identical to
those obtained in the analysis reported
sures. Fagot (1985) similarly found no differ-
ence between children at Levels 1 and 2 in above.

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 Bussey and Bandura 1243

0--- GIRLS
5 o- -o BOYS
q
z
0 O,
4-

SI I

YOUNGER OLDER

Masculine Feminine Masculine Feminine

G ENDER-LIN KED ACTIVITIES

GENDER-LINKED ACTIVITI ES

FIG. 1.-Mean self-evaluative reactions toward gender-linked behavior as a fun

of subject, and gender-linked activity.

Gender-Linked Conduct in which age was substituted for gender con-

stancy level.
Children's gender-linked behavior was
scored as the total number of seconds the
Cross-sex conduct.-A similar analysis
children played with each of the toys. Table
of variance was performed for children's
2 presents the means and standard devia- gender-linked behavior when only the cross-
tions for gender-linked behavior performed
sex toys (highly and moderately sex-typed)
by boys and girls at different levels of gen-
were available. The main effect for sex of
der constancy.
subject was significant, F(1,36) = 4.51, p <
These data were analyzed by a 2 (sex) .05, as was the main effect for cross-sex play
x 2 (gender constancy) x 2 (gender-linked material, F(1,36) = 16.98, p < .0001. The
behavior) analysis of variance with gender- interaction involving sex of subject and gen-
typed behavior as the within-subjects vari- der-linked activity was also significant,
F(1,36) = 5.72, p < .05. This interaction is
able. This analysis yielded a highly signifi-
cant interaction for sex of subject and depicted graphically in Figure 3. Boys and
gender-typed activity, F(1,36) = 31.04, p < girls did not differ in the extent of their mod-
.0001. This interaction is depicted graphi- erately cross-sex behavior; girls, however,
cally in Figure 2. engaged in significantly more highly cross-
sex behavior than did boys. As in the previ-
Boys performed the masculine activities ous analyses, there was no main effect for
for a longer period of time than the feminine
gender constancy, nor did it interact with
activities. Conversely, girls performed the any of the other variables. A similar analysis
feminine activities for a greater amount of
using age rather than gender constancy level
time than they did the masculine activities.
yielded the same pattern of results.
There was no main effect for gender con-
stancy, nor did it interact with any other fac- Relation of Theoretical Determinants to
tors. A comparable analysis using all four Gender-Linked Conduct
levels of gender conception yielded the A major issue addressed in this study
same pattern of results, as did the analysis is the relation of children's self-evaluative

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 1244 Child Development
TABLE 2

MEANS (and Standard Deviations) FOR GENDER-LINKED CONDUCT PERFORMED B

BOYS AND GIRLS AT DIFFERENT LEVELS OF GENDER CONSTANCY

GENDER CONSTANCY LEVEL

I II

GENDER-LINKED CONDUCT Girls Boys Girls Boys

All toys:
Feminine behavior ............................................................. 69.6 32.5 61.7 22.0
(70.0) (30.9) (50.4) (35.4)
Masculine behavior ..................................... 9.3 57.7 9.7 102.7
(12.7) (45.0) (16.8) (55.9)

Cross-sex toys:
High ........................................ 24.8 3.3 44.0 1.2
(24.1) (7.8) (26.9) (1.9)
Medium ........................................ 49.5 63.2 44.7 36.9
(35.7) (55.6) (26.0) (40.8)

reactions to their gender-linked in Table 3. For

conduct at the older children, self-
different age levels. To assess this self-reg-
evaluative reactions were consistent pre-
ulative function, children's gender-linked
dictors of gender-linked behavior. The
behavior was correlated with theirstronger their self-approval for feminine-
anticipa-
linked behavior,
tory self-evaluative reactions associated with the more they engaged
each gender-linked activity. These in traditionally
correla- feminine-typed activities
tions, as well as those for gender concep-
and refrained from traditionally masculine-
tions and gender knowledge, aretyped presented
activities. Conversely, the more

90

---- GIRLS

80 q o_--- BOYS
z
70- \
0
O

CC- 60-
w
o 50-
0
z 40- \

m 30-
LL..
o o
Cr
w 20-

10-

0
Masculine Feminine
GENDER- LINKED ACTIVITIES

FIG. 2.-Mean duration of gender-linked behavior as a function of sex of subject and gender-linked
activity.

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 Bussey and Bandura 1245

50-
5 P
----* GIRL
o--o BOYS

S40-
z
0

S30- o /
0

< 20/
<20

High Moderate
CROSS-SEX ACTIVITIES
FIG. 3.-Mean duration of cross-sex behavior as a function of sex of su

self-approving they were of masculine-

to self-evaluative reactions for same-sex be-
linked behavior, the more they havior
performed
(r = .14), but related to self-
masculine-typed activities and evaluative
shunned reactions for cross-sex behavior
feminine-typed activities. However, the
(r = -.43, p < .01). Thus, with increasing
younger children's reactions were age
unrelated
children expressed increasingly critical
to either feminine-linked or masculine- self-reactions for cross-sex behavior.
linked behavior.

To evaluate the relation between chil- Discussion

dren's gender conceptions and their gender-
The findings of this study lend suppor
linked behavior, their gender conception
scores were correlated with their duration of to a social cognitive model of gender dev
play with feminine-linked and masculine- opment as involving a shift from social
guided control to self-regulatory control o
linked materials (Table 3). Neither gender
gender-linked behavior with increasing age
identity, gender stability, level of gender
Younger children did not exhibit any diffe
constancy, nor gender-linked knowledge
ential anticipatory self-reactions to same-s
bore any relation to gender-linked conduct.
or cross-sex behavior, whereas older chi
Not surprisingly, gender identity, stability,
dren reacted self-approvingly for same-s
and constancy are highly interrelated.
behavior and self-critically for cross-sex b
Relation of Gender-Linked havior. This differential self-reaction was
Variables to Age true for both older boys and girls. Thus, with
A further set of correlations was com- increasing age children revealed a greater
puted relating age to level of gender concep- ability for anticipatory self-regulation of gen-
tion and gender-linked behavior. Age was der-linked behavior. Moreover, the older
significantly correlated with children's gen- children's anticipatory self-evaluative reac-
der-constancy score (r = .65, p < .0001) and tions predicted their actual gender-linked
their gender-knowledge score (r = .38, p < behavior. They engaged in the gender-
.01). However, age was unrelated to same- linked behavior they regarded self-
sex (r = .07) or opposite-sex (r = .04) gen- approvingly but shunned cross-sex behavior
der-linked behavior. Age was also unrelatedthat would lead them to react self-critically.

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 Zt Z
0

Cd 00 . - - P. 4C I ' .t
I I I I'l o I'l
Acl I1

0,-4 0 P- CI QC
O 40 * . 001 0 C-IO co in Q

?4 cl4-
00 -l

w00 0
z Q

0 *- *~-
* -(
00
C I
C c
z$ .
00

00 00

*C

6 dcd

P- . Q).W. C0
H -o

0 00 *CdCO
C3 C1 .
Ci2t b o
t~Cd Cd3
. . . 0 .
Z
0~ 0)
c3"~C "~ + + i
*c S ~- oo
o oz
c3 I 3 Cd Q) -4;jQ '
z x
S-4 z
0~ x L
O &3 ++ .s
.0
n n0

0 0- ~ o

- Q:

P-4 0 ~
n ~
-- ~ ~ d 0 Q )>~o H
"0 "0 "0 " 0 "0"
0
0)
00)
0I 0 002Sd0
0r o
0 0, 09 0 ~ E000

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 Bussey and Bandura 1247

Younger children neither exhibited any ing of the peer videotape was no easy matter,
differential anticipatory evaluative self-
either. As one 7-year-old boy remarked at
reactions nor any linkage between antic-
the completion of the filming, "It's the most
ipatory self-reactions and gender-linkedawful thing I have ever done." Although
conduct. Thus, with increasing age, chil-
girls expressed much weaker self-evaluative
dren's gender-linked behavior came increas-
reactions to cross-sex behavior, some of their
ingly under self-regulatory control. comments were most revealing. In express-
ing her self-sanctions against playing with a
With expanded gender-related experi-
ences, children's cognitive understanding of
truck, one girl explained, "My mommy
would want me to play with this, but I don't
gender emerges, as do personal standards for
want to." Her personal standards had evi-
gender-related conduct. Children showed a
dently come under the sway of extrafamilial
developmental change in which social sanc-
influences.
tioning of gender-related behavior precedes
self-sanctioning of the same activities. Even The findings do not support the view
the youngest children in this study behavedthat children were striving to match their be-
toward others in a gender-stereotypic man- havior to their gender labeling. From as
ner, despite their limited gender-linked
young as 30 months children chose to play
with same-sex toys. Their behavior con-
knowledge. Consistent with the social cog-
nitive theory of gender self-regulation, the
formed to gender-linked stereotypes regard-
findings of this study show that children first
less of level of gender conception. Children
learn to discriminate and evaluate gender- were not first labeling play material in a gen-
linked conduct and later to guide their ownder-linked manner and then engaging in
conduct by self-evaluative reactions. The
gender-linked behavior. Gender-linked be-
youngest children disapproved of peers'havior was guided by factors other than
cross-sex conduct but did not apply evalua-
matching gender labeling with behavior.
tive standards to their own gender-related
These findings offer little support for either
behavior. Irrespective of children being able
cognitive-developmental theory or for gen-
der schema accounts of gendered develop-
to label objects as gender-linked, they were
aware of the social standards associated with
ment. Although this study supported the
gender-linked objects. Most children, by 3finding of Martin and Little (1990) linking
years of age, would have been exposed to gender stability to gender knowledge, nei-
the play material used in this study and ther gender knowledge nor gender stability
would have learned which ones were ac- predicted gender-linked behavior. Martin
ceptable for them and for others to play with
and Little (1990) found that gender stability
and which ones were not (Caldera, Huston, was related to both children's gender knowl-
& O'Brien, 1989). edge and gender-linked preferences. It
When only cross-sex material was avail-
should be noted, however, that the present
study measured gender-linked behavior, not
able, boys displayed stronger self-sanctions
verbalized preferences. Gender preferences
against cross-sex behavior than did girls.
are not always predictive of gender-linked
Some tried to have the stereotypic feminine
toys removed. For example, when it became
conduct (Huston, 1983). Children's gender
apparent that they were being left with knowledge was highly correlated with age.
"feminine" toys, one boy hastily announcedConsistent with previous research (Bussey
& Bandura, 1984; Signorella, 1989), gender
to the departing experimenter, "No, I'm fin-
knowledge was unrelated to gender behav-
ished with those toys," even though he had
ior. This finding is not surprising since most
completely shunned them. They were not at
children even at an early age are fully aware
all hesitant in expressing their displeasure
with the selections they were left with:of"Ithe gender-linked stereotypes but show
don't like baby dolls." During the sessionsubstantial variation in gender-linked be-
havior.
they tried to do anything but play with the
cross-sex toys. One boy flung the doll across The relation between children's antici-
the room and turned his back on it, getting patory self-evaluative reactions for gender-
related behavior and their actual behavior is
it at least out of sight if not out of mind. Some
sought to restructure their limited options consistent with predictions from social cog-
by sticking to the moderately sex-typed ma- nitive theory. Awareness or knowledge of
terial and transforming it into masculine the gender linkages of objects and activities
tools, as, for example, using beaters in the alone will not produce behavior in line with
cooking set as guns or drills. Getting boysthe to stereotypes unless one is motivated to
change diapers on the baby doll for the film- conform to them. What is required is a mech-

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 1248 Child Developmrent
anism that links knowledge to action. The studies that chart the emergence and
present study provides supportive evidence changes in self-evaluative standards and
for a self-regulatory mechanism rooted in their impact on gender-linked behavior.
an internalized standard as an important
guiding and motivating link between gen- Despite the emphasis in recent years on
der knowledge and gender-linked conduct. gender egalitarianism, the children in this
These findings add to a growing literature study seem to be as stereotypically sex-
on the self-regulation of conduct by veri- typed as those of yesteryear. Where there
fying the operation of this mechanism in was a break away from the stereotypic pat-
children's gender development. tern, it was with girls rather than boys. A
number of previous studies have shown that
The findings taken as a whole reveal girls are less sex-typed than boys (Katz &
that from an early age children adopt tradi- Boswell, 1986) and that their sex-typed be-
tional patterns of gender-linked conduct. havior is more modifiable than that of boys
Neither gender constancy nor gender knowl- (Katz, 1986). It is hardly surprising that chil-
edge appear to guide this conduct. Rather, dren's gender-linked conduct continues to
children learn the social sanctions against conform to stereotypic gender-linked stan-
cross-sex behavior and social approval for sa- dards. Traditionally female-related activities
me-sex behavior and direct their own behav- and characteristics are still less valued than
ior accordingly. They, in turn, influence male-related ones (Connor & Serbin, 1978;
their peers by approving and disapproving Hall & Halberstadt, 1980; Zalk & Katz,
reactions to conform to the prevailing social 1978). There is little incentive for boys to
standards. Eventually, children adopt self- abandon the status and privilege accorded
evaluative standards for gender-linked be- the male role, unless females are portrayed
havior and regulate their own conduct as possessing the same status and benefits
through anticipatory self-sanctions. accorded males (Bandura, Ross, & Ross,
1963; Bussey & Bandura, 1984).
Given the influential role that self-
evaluative standards play in the self- Certain classes of behaviors and attri-
regulation of gender-linked conduct, the ap-butes, such as toy play and dress and hair
propriate next stage for research is to clarifystyles, convey highly salient and concrete
the processes by which children construct gender-related information that is easily re-
self-regulative standards from the diversealized even by very young children, so it is
sources of social influences that impingenot surprising that gender-linked knowledge
upon them. In this construction process they and conduct emerge in these aspects. Differ-
must select, weight, and integrate informa-ent clothing styles and activities remain gen-
tion from a variety of sources including par- der linked across the life span, but stereo-
ents, teachers, peers, and the electronic me-typic gender conceptions extend beyond
dia in forming their gender-linked standards these highly salient aspects to include a vast
(Bandura, 1986). It would be expected thatarray of human activities. Stereotypic gen-
as children's self-regulative functions de-
der-role socialization has reverberating ef-
velop, their gender-linked standards wouldfects into adulthood. A good case in point
have an increasing impact on their gender-are the psychosocial influences that shape
linked behavior. women's beliefs in their self-efficacy for dif-
ferent occupational pursuits (Bandura,
Gender-linked standards do not remain
1991c; Betz & Hackett, 1986; Matsui, Ikeda,
static. As children move increasingly into & Ohnishi, 1989). Women judge themselves
the larger community, they become aware of highly efficacious for occupations tradition-
the diversity of gender-linked practices. Not ally held by women but inefficacious to mas-
only is the range of their social experiences ter the demands of vocations dominated by
expanded, but broader social changes alter men, even though they have the ability for
the particular constellations of attributes that them. The instilled self-efficacy beliefs have
become linked to gender (Spence, 1985). In substantial impact on their career choice and
recent years, for example, long flowing locks development (Lent & Hackett, 1987).
and cooking skills have become compatible
with masculinity. The developmental course The marked sex segregation that occurs
that self-regulative standards take is thus an- from about 3 years of age onward contributes
other issue of considerable interest. The re- importantly to strong adherence to stereo-
lations obtained in the present study aretypic gender-linked standards (Maccoby,
based on cross-sectional data. They need to1990). However, children whose parents es-
be verified in experimental and longitudinal pouse and model gender egalitarianism,

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 Bussey and Bandura 1249

who are in educational systems that foster and secondary reinforcement theories of
such standards, and who interact with peers identificatory learning. Journal of Abnormal
who endorse them are more apt to develop and Social Psychology, 67, 527-534.
egalitarian gender-linked standards. Unlike
Bem, S. L. (1981). Gender schema theory: A cogni
moral standards, where there are incentives tive account of sex typing. Psychological Re-
for selective engagement of self-regulative view, 88, 354-364.
control (Bandura, 1986, 1991a), there are Betz,
no N. E., & Hackett, G. (1986). Applications
parallel incentives for disengaging stereo- of self-efficacy theory to understanding career
typic gender-linked standards. Pressure choice behavior. Journal of Social and Clini-
from parents alone is unlikely to achieve cal Psychology, 4, 279-289.
egalitarian standards and conduct in chil-Blakemore, J. E. O., LaRue, A. A., & Olejnik,
dren unless the peer group and other sig- A. B. (1979). Sex appropriate toy preference
nificant elements in the culture endorse and the ability to conceptualize toys as sex-
them (Weisner & Wilson-Mitchell, 1990). role related. Developmental Psychology, 15,
Because of the overwhelming sex segrega-339-340.
tion of children and strong pressures forBussey, con- K., & Bandura, A. (1984). Gender con-
formity from the peer group, stereotypic stancy, social power, and sex-linked model-
gender-linked standards may be particularly ing. Journal of Personality and Social Psy-
stable and resistant to modification in the ab- chology, 47, 1242-1302.
sence of sweeping social changes. Caldera, Y. M., Huston, A. C., & O'Brien, M.
(1989). Social interactions and play patterns
Social cognitive theory accords an in-
fluential role to institutional structures and of parents and toddlers with feminine, mascu-
line, and neutral toys. Child Development,
sanctions in shaping gender roles. However,
structural influences on individual behavior 60, 70-76.
Carter, D. B., & Levy, G. D. (1988). Cognitive as-
are mediated, in large part, through self-
pects of children's early sex-role develop-
referent processes (Bandura, 1991b). This
ment: The influence of gender schemas on
causal structure requires integration of so-
preschoolers' memories and preferences for
ciocultural determinants with personal de-
sex-typed toys and activities. Child Develop-
terminants. Studies that relate changing so-
ment, 59, 782-793.
cietal valuation of gender-linked activities
Connor, J. M., & Serbin, L. A. (1978). Children's
to children's construction of personal stan-
responses to stories with male and female
dards hold considerable promise in further-
characters. Sex Roles, 4, 637-645.
ing our understanding of the self-regulation
Edwards, V. J., & Spence, J. T. (1987). Gender-
of gender-role development.
related traits, stereotypes, and schemata.
References Journal of Personality and Social Psychology,
53, 146-154.
Fagot, B. I. (1985). Changes in thinking about
Bandura, A. (1986). Social foundations of thought
and action: A social cognitive theory. Engle-early sex role development. Developmental
wood Cliffs, NJ: Prentice-Hall. Review, 5, 83-98.
Bandura, A. (1989). Social cognitive theory. In R.B. I., & Leinbach, M. D. (1989). The young
Fagot,
Vasta (Ed.), Annals of child development: child's gender schema: Environmental input,
Vol. 6. Six theories of child development (pp.
internal organization. Child Development, 60,
1-60). Greenwich, CT: JAI. 663-672.
Bandura, A. (1991a). Social cognitive theory Fagot,ofB. I., Leinbach, M. D., & Hagan, R. (1986).
moral thought and action. In W. M. Kurtines Gender labeling and the adoption of sex-
& J. L. Gewirtz (Eds.), Handbook of moral typed behaviors. Developmental Psychology,
behavior and development (Vol. 1, pp. 45- 22, 440-443.
103). Hillsdale, NJ: Erlbaum. Grusec, J. E., & Kuczynski, L. (1977). Teaching
Bandura, A. (1991b). Self-regulation of motivation
children to punish themselves and effects on
through anticipatory and self-regulatory subsequent compliance. Child Development,
mechanisms. In R. A. Dienstbier (Ed.), Per- 48, 1296-1300.
spectives on motivation: Nebraska Sympo- Hall, J. A., & Halberstadt, A. G. (1980). Masculin-
sium on Motivation (Vol. 38, pp. 69-164).ity and femininity in children: Development
Lincoln: University of Nebraska Press. of the Children's Personal Attributes Ques-
Bandura, A. (1991c). Self-efficacy: The exercisetionnaire.
of Developmental Psychology, 16,
control. Unpublished manuscript, Stanford 270-280.
University, Stanford, CA. Huston, A. C. (1983). Sex typing. In E. M. Hether-
Bandura, A., Ross, D., & Ross, S. A. (1963). A com-
ington (Ed.), P. H. Mussen (Series Ed.),
parative test of the status envy, social power,
Handbook of child psychology. Socialization,

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms
 1250 Child Development

personality, and social development (Vol. Rheingold,

4, H. L., & Cook, K. V. (1975). The con-
pp. 387-467). New York: Wiley. tent of boys' and girls' rooms as an index of
Katz, P. A. (1986). Modification of children's gen- parents' behavior. Child Development, 46,
der-stereotyped behavior: General issues and 459-463.
research considerations. Sex Roles, 14, 591- Signorella, M. L. (1987). Gender schemata: In-
602. dividual differences and context effects. In
Katz, P. A., & Boswell, S. (1986). Flexibility and L. S. Liben & M. L. Signorella (Eds.), Chil-
traditionality in children's roles. Genetic, So- dren's gender schemata (New Directions for
cial, and General Psychology Monographs, Child Development, No. 38, pp. 23-37). San
112, 105-147. Francisco: Jossey-Bass.
Kohlberg, L. (1966). A cognitive-developmental Signorella, M. L. (1989, April). Gender-role atti-
analysis of children's sex-role concepts and tudes versus knowledge in the study of chil-
attitudes. In E. E. Maccoby (Ed.), The devel- dren's gender schemas. Paper presented at
opment of sex differences (pp. 82-173). Stan- the meeting of the Society for Research in
ford, CA: Stanford University Press. Child Development, Kansas City.
Lamb, M. E., & Roopnarine, J. L. (1979). Peer in- Slaby, R. G., & Frey, K. S. (1975). Development
fluences on sex-role development in pre- of gender constancy and selective attention
schoolers. Child Development, 50, 1219- to same-sex models. Child Development, 46,
1222. 849-856.
Lent, R. W., & Hackett, G. (1987). Career self- Smetana, J. G., & Letourneau, K. J. (1984). Devel-
efficacy: Empirical status and future direc- opment of gender constancy and children's
tions. Journal of Vocational Behavior, 30, sex-typed free play behavior. Developmental
347-382. Psychology, 20, 691-696.
Maccoby, E. E. (1990). Gender and relationships:Spence, J. T. (1985). Gender identity and its im-
A developmental account. American Psychol- plications for the concepts of masculinity and
ogist, 45, 513-520. femininity. In T. B. Sonderegger (Ed.), Ne-
Marcus, D. E., & Overton, W. F. (1978). The de- braska Symposium on Motivation (Vol. 32,
velopment of cognitive gender constancy and pp. 59-95). Lincoln: University of Nebraska
sex role preferences. Child Development, 49, Press.
434-444.
Stangor, C., & Ruble, D. N. (1987). Development
Martin, C. L., & Halverson, C. F. (1981). A sche- of gender role knowledge and gender con-
matic processing model of sex typing and ste-
stancy. In L. S. Liben & M. L. Signorella
reotyping in children. Child Development,
(Eds.), Children's gender schemata (pp. 5-
52, 1119-1134. 22). San Francisco: Jossey-Bass.
Martin, C. L., & Little, J. K. (1990). The relation of
Weinraub, M., Clemens, L. P., Sockloff, A.,
gender understanding to children's sex-typed Ethridge, T., Gracely, E., & Myers, B. (1984).
preferences and gender stereotypes. Child The development of sex role stereotypes in
Development, 61, 1427-1439. the third year: Relationships to gender label-
Matsui, T., Ikeda, H., & Ohnishi, R. (1989). Rela-
ing, gender identity, sex-typed toy prefer-
tions of sex-typed socializations to career self-
ences and family characteristics. Child Devel-
efficacy expectations of college students. opment, 55, 1493-1503.
Journal of Vocational Behavior, 35, 1-16. Weisner, T. S., & Wilson-Mitchell, J. E. (1990).
Perry, D. G., & Bussey, K. (1977). Self- Nonconventional family life-styles and sex
reinforcement in high- and low-aggressivetyping in six-year-olds. Child Development,
boys following acts of aggression. Child De-61, 1915-1933.
velopment, 48, 653-658. Zalk, S. R., & Katz, P. A. (1978). Gender attitudes
Perry, D. G., Perry, L. C., Bussey, K., English, D., &
in children. Sex Roles, 4, 349-357.
Arnold, G. (1980). Processes of attributionZimmerman,
and B. J. (1989). A social cognitive view
children's self-punishment following misbehav- of self-regulated academic learning. Journal
ior. Child Development, 51, 545-551. of Educational Psychology, 81, 329-339.

This content downloaded from 193.144.196.28 on Tue, 28 Aug 2018 10:00:50 UTC
All use subject to https://about.jstor.org/terms