Bartsch2018 (Personalized Charity Ad)
Bartsch2018 (Personalized Charity Ad)
To cite this article: Anne Bartsch & Andrea Kloß (2019) Personalized charity advertising. Can
personalized prosocial messages promote empathy, attitude change, and helping intentions
toward stigmatized social groups?, International Journal of Advertising, 38:3, 345-363, DOI:
10.1080/02650487.2018.1482098
Introduction
Prosocial media messages such as charity appeals, public awareness campaigns, or
messages to combat prejudice and discrimination of stigmatized social groups provide
an important avenue for promoting altruism and social cohesion in today's mass soci-
eties. For example, charity advertising can help raise donations to alleviate the suffer-
ing of those affected by illness, war, or poverty (Bennett 2015; Chang and Lee 2010;
Griffin et al. 1993; Kim 2014). Prosocial messages can also raise public awareness about
social, environmental, and health issues (Bagozzi and Moore 1994; Mares and
Woodard 2005; Shen 2011), or they can help transform individuals’ attitudes and
behaviour toward members of stigmatized groups such as ethnic minorities, persons
with illnesses or disabilities, and poor, elderly or homeless people (Batson et al. 2002;
Oliver et al. 2012).
With the proliferation of media channels and available content, however, prosocial
messages are competing for an increasingly scarce resource, that is, individuals' atten-
tion and willingness to seriously consider the message. Before a prosocial message
can have its intended effect, it first needs to get the audience involved; in addition,
the message needs to persuade recipients, which is not an automatic byproduct of
message exposure (Petty and Cacioppo 1986). Several lines of research have identified
characteristics of prosocial messages that are helpful in promoting this dual goal of
message involvement and persuasion. For example, research on charity advertising
has examined the effects of message factors such as emotional appeals (Basil, Ridgway,
and Basil 2006; Bennett 2015; Chang 2011; Small and Verrochi 2009), message framing
in terms of negative consequences of not helping vs. positive consequences of help-
ing (Chang and Lee 2010; Das, Kerkhof, and Kuiper 2008), altruistic vs. egoistic appeals
(Chang 2014; Chang and Lee 2011; Kim 2014), and vividness of case stories vs. statistical
evidence (Chang and Lee 2010; Das, Kerkhof, and Kuiper 2008). Similar factors have
also been studied with regard to other types of prosocial messages, for example in
the context of public service announcements, health communication, journalism, and
political communication (Oliver et al. 2012; Shen 2011).
The present study aims to extend research on charity advertising with a special
focus on message personalization as an additional content feature that might foster
both audience involvement, and persuasive effects. Digital media technologies have
made it easier to personalize messages. For example, the recipient can be addressed
by name, or the message can be customized to the recipient's preferences based on
information contained in user profiles (Kalyanaraman and Sundar 2006). In addition,
the viral dissemination of messages through social networks allows for a personalized
distribution of messages where the sender is personally known to the recipient
(Petrescu and Korgaonkar 2011).
Personalized features are increasingly being used by charity campaigns and advo-
cacy groups for prosocial purposes but their effects are not sufficiently understood.
The available research on personalized messages has mainly focused on product
advertising so far, whereas research on personalized features in charity advertising has
remained scarce (Masthoff, Langrial, and van Deemter 2013). It further remains unclear
whether personalized prosocial messages can have a downside too. Research on per-
sonalized advertisements has found that ‘getting too personal’ (White et al. 2008) can
backfire in that it produces reactance – that is a sense of threatened freedom of opin-
ion and actions that can lead to anger, counterarguing, and message rejection (Dillard
and Shen 2005; Rains 2013; Shen 2011). Hence, in addition to exploring prosocial
effects of personalization, it also seems important to consider the possibility that recip-
ients of personalized charity advertising might feel pressurized and might respond
with reactance. To better understand the effects of personalized charity advertising,
the research literature on personalized media messages in other domains such as
product advertising and political communication is reviewed and linked to research on
prosocial media messages in general and to research on charity advertising
in particular.
Based on this combined theoretical framework, we developed a set of hypotheses
that are tested in an online experiment using a personalized and a nonpersonalized
INTERNATIONAL JOURNAL OF ADVERTISING 347
version of a charity advertisement for the Big Boston Warm-Up campaign that solicits
donations of winter coats for homeless people and aims to promote public awareness
and attitude change about this stigmatized group. Specifically, we expected that, in
the context of charity advertising, message personalization could have positive as well
as negative effects on prosocial outcomes such as changes in attitudes and behav-
ioural intentions toward the stigmatized group, and willingness to support a donation
campaign. On the one hand, we assumed that personalization can reinforce prosocial
outcomes (H1), and that this positive effect is mediated by involvement (H2), and by
empathy (H3). On the other hand, we considered the possibility that personalization
may elicit reactance which could act as a negative mediator that runs counter to these
positive effects (H4). Finally, we examined whether involvement can mitigate reactance
and its negative effects (H5). The combined effects assumed in this set of hypotheses
are represented in the form of a path model in Figure 1. The following section
explains the theoretical rationale behind each of the hypotheses and reviews research
findings from other domains such as product advertising, public service announce-
ments and political communication that informed our predictions. We, then, report
the methods and results of an experiment to test these assumptions and discuss the-
oretical and practical implications of our findings.
Theoretical background
Can personalized messages be harnessed for prosocial purposes such as transforming
audiences' attitudes and behaviour toward stigmatized social groups, or raising charit-
able donations? To answer this question, it first needs to be clarified what personaliza-
tion means. According to Vesanen’s (2007) analysis, personalization is a multifaceted
concept including different forms of customization of products and services as well as
personalized communication with the customer. Our focus is on the latter aspect, per-
sonalized communication, which is described by White et al. (2008, 40) as ‘specialized
flow of communication that sends different recipients distinct messages tailored to
their individual preferences or characteristics’ (see also Kalyanaraman and Sundar
2006). For example, messages can be personalized by addressing the message recipi-
ent by name, or by including other information assumed to be relevant to the recipi-
ent based on his or her user preferences, interests, or purchase history.
Generally, there is reason to believe that the personalization of messages is condu-
cive to positive, message consistent effects. For example, research has found that
348 A. BARTSCH AND A. KLOß
involvement (Johnson and Eagly 1989). Research on the effects of audiences’ involve-
ment with narrative entertainment has distinguished between perceived similarity, per-
spective-taking, parasocial interaction, and transportation (Tukachinsky and Stokunaga
2013). In the context of advertising, involvement has been conceptualized in terms of
attention/processing strategies, personal/situational involvement, audience/process
involvement, and enduring/product involvement (Andrews, Durvasula, and
Akhter 1990).
The dimension of involvement most pertinent to our research on personalized char-
ity advertising is self-referencing, ‘the processing of information by relating it to the
self-structure or aspects of it’ (Burnkrant and Unnava 1995, 17). This form of involve-
ment is characterized by cognitions that link the message to recipients’ self-concept in
a direct way, whereas other forms of involvement relate to the self indirectly (e.g.
through personal relevance of issues, goals, values, products, and so forth, or through
identification with characters in a story). Despite its specific focus, self-referencing is
similar to other forms of involvement in that it shares the common conceptual core of
personal relevance (Petty and Cacioppo 1986), and a similar pattern of persuasive
effects (Burnkrant and Unnava 1989, 1995).
The focus of our research on self-referencing is not meant to imply that the effects
of message personalization are limited to this particular form of involvement. For
example, charity advertising can be personalized with regard to other dimensions of
involvement such as perceived similarity with victims (Masthoff, Langrial, and van
Deemter 2013), or personal world-views and values (Brunel and Nelson 2000; Chang
and Lee 2011). However, given the lack of research on direct forms of personalization
in charity advertising (such as addressing recipients by name), we focused on self-
referencing as the corresponding dimension of involvement that should serve as a crit-
ical mediator of persuasive effects in the case of personalized messages that directly
address the recipient’s self. Specifically, we expected that the effect of personalized
charity advertising on prosocial outcomes assumed in H1 is mediated by self-referenc-
ing (H2).
always backfire in that it creates reactance (Dillard and Shen 2005; Maslowska 2011;
White et al. 2008). Reactance is defined as a ‘motivational state directed toward the
reestablishment of threatened or eliminated freedom’ (Brehm 1966, 15). In the case of
persuasive messages, recipients are often aware of the persuasive intent of the mes-
sage and may feel that the message threatens their freedom of opinion and action.
The motivational state of reactance aimed at the reestablishment of their threatened
freedom can lead recipients to react with anger, counterarguments, and with attitudes
and behaviour that run counter to the message intent (Dillard and Shen 2005; Rains
2013). Generally, the inclusion of personalized content may heighten individuals'
awareness of persuasive message intent and may thereby elicit reactance. For
example, in a study of personalized advertising, White et al. (2008) found that highly
personalized e-mails produced reactance and lowered recipients' willingness to con-
sider the offer when the perceived utility of the offer was low. Likewise, research of
Maslowska (2011) suggests that personalization can be counterproductive if recipients
have low trust in the company, or high privacy concerns.
Prosocial messages such as charity advertisements that aim to elicit altruistic
thoughts, feelings, and behaviours are particularly vulnerable to reactance effects. As
Weinstein and Ryan (2010) have argued, a sense of external pressure and reactance
can be detrimental to individuals' altruistic motivation. Based on self-determination
theory (Ryan and Deci 2000), these authors predicted and found that the psycho-
logical benefits of helping were contingent on individuals’ autonomous choice to per-
form the helping behaviour. Under conditions of external control or moral pressure,
the helping experience was less than gratifying. In the context of charity advertising,
personalization might therefore be a double-edged sword because it can raise recipi-
ents’ perception of persuasive intent and normative pressure to a degree that hurts
their sense of autonomy, which in turn elicits reactance and undermines their altruistic
motivation. Thus, it can be assumed that personalization of charity advertising
increases reactance, thereby reducing prosocial outcomes (H4).
Conversely, the self-reference effect of personalization might (partly) offset react-
ance. If a personalized message successfully invites recipients to get involved through
self-referencing, then their tendency to feel pressured and to develop reactance
should be diminished. Indirect evidence for this assumption comes from research on
narrative persuasion (Green and Brock 2000; Slater and Rouner 2002). This line of
research has found that the more individuals are involved and absorbed in a story the
less likely they are to report critical counterarguments against the persuasive message
of the narrative. Several authors have argued that self-referencing plays an important
role in narrative involvement (Cupchik 1995; Wirth, Hofer, and Schramm 2012). In add-
ition to character empathy and perceived realism, audiences’ ability to relate the story
to their own life and experiences constitutes a critical involvement factor. Hence, in
the case of narrative messages, personalization may reinforce the self-referencing facet
of narrative involvement by encouraging recipients to relate the story to their self-con-
cept and personal experiences, which in turn should reduce reactance. The role of nar-
rative involvement in mitigating reactance has only been tested with regard to the
broader concept so far. Nevertheless, it seems reasonable to assume that the specific
dimension of involvement considered in our study (self-referencing) would show the
352 A. BARTSCH AND A. KLOß
same mitigation effect. Therefore, we assumed that (in addition to the mediating role
of empathy) the effect of self-referencing on prosocial outcomes predicted in H2 is
mediated by reduced reactance (H5).
Figure 1 summarizes the hypotheses and research question in the form of a path
model where the effect of personalization on prosocial outcomes (H1) is mediated by
self-referencing (H2), and the effect of self-referencing on prosocial outcomes is medi-
ated by increased empathy (H3), and reduced reactance (H5). In addition, the model
includes a negative direct effect of personalization on reactance (H4). In this model,
prosocial outcomes are treated as a latent variable that is estimated from individuals'
reported attitudes and behavioural intentions toward the stigmatized group, and from
their reported willingness to give to a charity campaign to help the stigma-
tized group.
Method
Overview
To examine our hypotheses, we conducted an online experiment with message per-
sonalization as the experimental factor. Personalization was operationalized using a
personalized and a nonpersonalized version of a charity advertisement for the Big
Boston Warm-Up campaign that solicits donations of winter coats for homeless people
and aims to promote public awareness and attitude change about this stigmatized
group. One version of the spot was personalized by including the recipient's name,
whereas the other version was nonpersonalized. Participants were randomly assigned
to see either the personalized or the nonpersonalized version of the spot and were
asked to rate their experience of self-referencing, their empathic and reactant
responses, as well as their attitudes and behavioural intentions toward homeless peo-
ple, and their willingness to support the campaign.
Stimuli
The original campaign was run by The Big Boston Warm-Up, an initiative to collect a
winter coat for every homeless person in Boston (for more informations about the
campaign see: http://lands-end-big-boston-warmup.archive.firstborn.com/). The spots
were produced in 2009 for both TV broadcast and viral distribution through social
media. The story told in both spots was identical: A young man unpacks a red coat
which he received as a Christmas present from his parents. The years are passing as
the young man goes to college, meets his wife, hugs his daughter, always wearing the
red coat. Finally, the coat ends up unused hanging on the hat stand. Then it is
donated for The Big Boston Warm-Up campaign and is handed to a homeless man
who smiles and strokes the coat. This part of the spot was identical in both versions.
The personalized version of the spot for viral distribution includes an additional short
scene that shows a homeless woman who receives a coat and looks at the label
inside. The name of the message recipient was displayed on the label, based on infor-
mation entered by the person who forwarded the spot in the viral campaign. Finally, a
text message was shown requesting the recipient to donate a winter coat at Land’s
INTERNATIONAL JOURNAL OF ADVERTISING 353
End Shop. In Return, a discount of 20% on a new Land’s End coat was offered. Again,
the text message addressed the recipient by name. In the nonpersonalized version
broadcast on TV, a voice-over message informed the viewer about the campaign after
the scene with the homeless man receiving the red coat. The spoken text was identi-
cal to the written text of the viral spot, except that the recipient was not addressed
by name. The two versions differed slightly in length. The nonpersonalized TV spot
was 1:00 min long, and the personalized viral spot was 1:20 min long.
Measures
Demographics
The first section of the questionnaire asked participants to provide basic demographic
information (age, gender, and education).
Self-referencing
Self-referencing was assessed with three items adapted from Burnkrant and Unnava
(1989): ‘the message seemed to be written with me in mind’, ‘the message seemed to
relate to me personally’, ‘the message had nothing to do with me’ (reverse scored;
a ¼ .70). A fourth item, ‘the message made me think about my personal experiences
354 A. BARTSCH AND A. KLOß
with donations’ lowered the reliability of the scale and was therefore excluded.
Ratings were recorded on a 7-point scale (1 ¼ do not agree at all, 7 ¼ fully agree).
Empathy
In line with Batson et al.’s (1997) definition, empathy was assessed as a state variable
(rather than a personality trait). We used four items from Batson (1987): ‘moved’,
‘tender’, ‘softhearted’, and ‘compassionate’ (a ¼ .90). Similar items have been used as
measures of sympathy (Eisenberg et al. 1991), warm feelings (Escalas, Moore, and
Britton 2004), and being moved (Bartsch and Schneider 2014). Ratings were recorded
on a 7-point scale (1 ¼ do not agree at all, 7 ¼ totally agree).
Figure 2. Structural equation model including standardized estimates of direct effects and variance
explained, direct paths that were nonsignificant are toned down.
as separate variables. Therefore, although both variables deal with behavioural inten-
tions, we analyzed them as separate variables to conserve the conceptual distinction
between behavioural intentions directly related to the campaign advertised in the
message, and behavioural intentions concerning the general issue of homelessness
beyond the specific content of the campaign.
Reactance
Reactance was assessed using Hastall and Sukalla's (2012) 5-item measure (e.g. ‘After
viewing the ad I would rather do the opposite of what was recommended in the ad’,
‘It frustrates me that the ad dictates me what to do’). Ratings were recorded on a 7-
point scale (1 ¼ do not agree at all, 7 ¼ totally agree, a ¼ .83). One item, ‘I consider rec-
ommendations in advertisements as an intrusion into my privacy’, lowered the reliabil-
ity of the scale, and was therefore excluded.
Results
To test our hypotheses, a structural equation model was calculated using AMOS.
Significance tests for direct, indirect, and total effects were estimated using bootstrap-
ping procedures employing 2000 bootstrap samples with a 95% bias-corrected confi-
dence interval. The analysis of the initial structural equation model revealed a good fit
to the data (v2 ¼ 13.04, df ¼ 9, p ¼ .16, CFI ¼ .98, RMSEA ¼ .055, 90% CI [.00–.12]). For
estimates of direct effects and variance explained, see Figure 2.
Hypothesis 1 assumed that the personalized stimulus spot would elicit a higher
level of prosocial outcomes including more positive attitudes and behavioural inten-
tions toward homeless people, and greater willingness to donate to the campaign.
The total effect of personalization on prosocial outcomes (b ¼ –.09, p > .05) was not
significant, and the direct path was marginally significant in the opposite direction
(b ¼ –.19, p ¼ .05). Thus, Hypothesis 1, the general assumption that personalization
would produce positive, message consistent effects, was not supported.
Hypothesis 2 made a more specific prediction by assuming that personalization
would have a positive effect on prosocial outcomes to the extent that it prompted
self-referencing. The personalized ad had a significant total effect on self-referencing
(b ¼ .32, p < .001), and self-referencing had a significant total effect on prosocial
356 A. BARTSCH AND A. KLOß
outcomes (b ¼ .23, p < .05). The indirect effect of personalization on prosocial out-
comes via self-referencing was significant as well (b ¼ .07, p < .01). Thus, despite the
missing direct link between personalization and prosocial outcomes, the mediation
hypothesis for self-referencing (H2) was supported. According to Hayes (2009), signifi-
cant indirect effects can be valid even in the absence of a significant direct or total
effect. This is because nonsignificant total effects can involve a confound of positive
and negative indirect effects that are mediated by different mediator variables. Such a
combination of positive and negative mediation effects can result in a nonsignificant
total effect even though individually the mediation effects are significant. As reported
below, a negative mediation effect of reactance (H4) was responsible for the nonsigni-
ficant total effect of personalization on prosocial outcomes in the case of our study.
Hypothesis 3 assumed that the effect of self-referencing on prosocial outcomes
would be mediated by increased empathy. Self-referencing had a significant direct
effect on empathy (b ¼ .44, p < .001), and empathy had a significant direct effect on
prosocial outcomes (b ¼ .20, p < .05). The indirect effect of self-referencing on prosocial
outcomes via empathy was significant as well (b ¼ .09, p < .05). Thus, the mediation
hypothesis for empathy (H3) was supported. A follow-up analysis revealed that person-
alization had a significant total effect on empathy (b ¼ .22, p < .01), and that this effect
was fully mediated by an indirect effect via self-referencing (b ¼ .14, p < .001), such
that the direct effect was reduced to nonsignificance (b ¼ .08, p > .05). This follow-up
analysis indicates a ‘clean’ manipulation of self-reference that was not confounded by
other factors that might have influenced empathy directly. Taken together, the medi-
ation analyses for self-referencing and empathy suggest the following causal sequence
of effects: (1) message personalization had a direct effect on self-referencing; (2) the
effect of personalization on empathy occurred indirectly, because self-referencing rein-
forced empathy; (3) self-referencing and empathy jointly contributed to prosocial
changes in attitudes, general behavioural intentions, and willingness to support
the campaign.
Hypothesis 4 assumed that personalization of prosocial messages increases react-
ance, thereby reducing prosocial outcomes. Personalization had a significant direct
effect on reactance (b ¼ .18, p < .05), and reactance had a significant negative effect
on prosocial outcomes (b ¼ –.20, p < .05). The indirect effect of personalization on pro-
social outcomes via reactance was negative and significant (b ¼ –.04, p < .05). Thus,
the negative mediation hypothesis for reactance (H4) was supported.
Hypothesis 5 assumed that the negative effect of reactance would be partly offset,
because self-referencing would lower reactance. As expected, self-referencing had a
significant negative effect on reactance (b ¼ –.25, p < .01), and reactance had a signifi-
cant negative effect on prosocial outcomes (b ¼ –.20, p < .05). The indirect effect of
self-referencing on prosocial outcomes via reactance was positive and significant
(b ¼ .05, p < .05). Thus, the hypothesis that reduced reactance would mediate the
effect of self-referencing on prosocial outcomes (H5) was supported.
In sum, although our hypotheses concerning the positive mediation effects of self-
referencing and empathy were supported, the negative mediation effect of reactance
suppressed these positive effects and led to a nonsignificant total effect of personal-
ization on prosocial outcomes.
INTERNATIONAL JOURNAL OF ADVERTISING 357
Discussion
This study aimed to investigate the role of personalized charity advertising in promot-
ing self-referencing, empathy, attitude change, and helping intentions toward stigma-
tized social groups. Based on theories of message involvement, empathy, and
reactance, we predicted that the personalization of charity advertisements would have
positive as well as negative effects on prosocial outcomes such as changes in attitudes
and behavioural intentions toward homeless persons and willingness to donate to the
campaign. On the positive side, personalization can reinforce self-referencing and
empathy, and can thereby promote prosocial outcomes. On the negative side, person-
alization can also elicit reactance, with negative effects on prosocial outcomes.
These assumptions were tested in an online experiment using a personalized and a
nonpersonalized version of an advertisement for The Big Boston Warm-Up campaign
that solicits donations of winter coats for homeless people and aims to promote pub-
lic awareness and attitude change about this stigmatized group. Contrary to the gen-
eral hypothesis that personalization would lead to message consistent effects,
personalization showed no significant direct effect on prosocial outcomes (H1).
Nevertheless, the results supported our assumption that self-referencing (H2) and
empathy (H3) are important mediating factors that contribute to positive effects of
personalized charity advertisements on prosocial outcomes. These positive effects
were qualified, however, by a direct effect of personalization on reactance (H4), which
suggests that for some individuals personalized charity advertisements might raise the
perception of persuasive intent and moral pressure to a degree that might suppress
the otherwise positive effects of personalization.
Reactance to personalized advertisements is a relatively new research topic but the
available findings (Maslowska 2011; White et al. 2008) suggest that personalized com-
munication can be a double-edged sword that can lead to reactance and boomerang
effects in certain situations (low perceived utility of the offer), and for certain groups
of recipients (those with low trust in the company and high privacy concerns). As
White et al. (2008) and Maslowska (2011) have argued, negative effects associated
with reactance might account for some of the weak or inconsistent findings concern-
ing the effects of personalization on message acceptance and persuasion.
To our knowledge, reactance has not been examined in the context of personalized
charity advertising so far. Therefore, our findings need to be interpreted with caution.
It might be the case that the personalized message was perceived by some partici-
pants as an attempt to exert moral pressure – which might explain their reactance
and reduced prosocial response tendencies. As Weinstein and Ryan (2010) have
argued, a sense of external pressure can be detrimental to individuals' altruistic motiv-
ation. Therefore, it seems important to keep potential drawbacks of message personal-
ization in mind – both with regard to personalized communication in general
(Maslowska 2011; White et al. 2008), and with regard to personalized charity advertis-
ing in particular.
On a more optimistic note, the results of our study suggest that the direct effect of
personalization on reactance (H4) was counterbalanced by an indirect effect via self-
referencing, such that the personalized message elicited greater self-referencing, which
in turn reduced reactance (H5). Combined, the positive and negative effects of
358 A. BARTSCH AND A. KLOß
personalization on reactance canceled each other out. Considering that the balance
between positive and negative effects of personalization on reactance might be
shifted toward either side depending on context factors, it would be premature to
draw firm conclusions.
What seems relatively clear, however, is that personalized charity advertising can
reinforce empathy. Examining this previously untested assumption was one of the pri-
mary aims of our study. The results indicate that personalization reinforced empathy,
as indicated by a significant total effect of personalization on empathy. The structural
equation model also provided insight into the reasons why empathy was reinforced
by message personalization, given that the effect of personalization on empathy was
mediated by self-reference. A possible explanation is that self-referencing helps miti-
gate implicit bystander effects resulting from the imagined presence of others who
could help instead (Garcia et al. 2002). The concept of an implicit bystander effect
aptly describes how diffusion of responsibility may occur among recipients of charity
advertisements who are addressed as part of a mass audience. Such implicit bystander
effects may be circumvented by a more personal form of address that can reinforce
recipients' sense of personal responsibility, thus increasing empathy and altruistic
motivation. To further substantiate this theoretical explanation, more research is
needed examining the causal role of the bystander effect. For example, experimental
manipulations could address message recipients' level of awareness that thousands of
others received similar personalized messages. If mitigation of the bystander effect
does in fact play a causal role, then such an awareness manipulation should disrupt
the effect of personalization on empathy.
As already mentioned, however, personalization can also be counterproductive to
the extent that it produces reactance. In sum, the positive effects of self-referencing
and empathy, and the negative effect of reactance resulted in a nonsignificant total
effect of personalization on prosocial outcomes, indicating that statistically the positive
and negative effects tended to cancel each other out.
In terms of practical implications, this complex pattern of findings including positive
as well as negative effects of personalized prosocial messages might appear somewhat
disappointing. To those who want to know whether it is a good or bad idea to per-
sonalize charity advertisements, the answer is: well, it depends. Although this might
not be the most satisfying answer, it is important to note the difference between ‘it
depends’ and ‘it doesn't matter’. A nonsignificant total effect does not necessarily
imply that for all recipients the personalized message was just as effective as the
unpersonalized one. It seems that for those participants who responded with self-
referencing and empathy, the personalized message was more effective, whereas for
those who responded with reactance it was less effective than the unpersonalized
message. Thus, a nonsignificant total effect can mean that in fact personalization did
matter, but that it mattered in different ways for different groups of recipients.
Therefore, an important next step is to examine the personal predispositions that
make individuals more or less likely to respond with self-referencing and empathy, or
with reactance to personalized prosocial messages. For example, examining the mod-
erating influence of personality traits such as trait empathy (Mehrabian and Epstein
1972), or reactance proneness (Hong and Faedda 1996) could inform our
INTERNATIONAL JOURNAL OF ADVERTISING 359
empathy occurred. The empathy effect of the personalized spot was fully mediated by
self-referencing. If confounding factors of the manipulation (such as length or gender
of the added character) had produced an empathy effect, it would have been
observed in the form of a direct effect of the manipulation on empathy which was
not the case. Finally, the assessment of prosocial outcomes was limited by the reliance
on self-report measures. These self-report data need to be complemented with implicit
attitude measures and behavioural measures to substantiate the current results.
With these limitations in mind, we think that the findings provide new insights into
the ‘black box’ (Kreuter et al. 1999) of personalization by highlighting the role of medi-
ating variables such as self-referencing, empathy, and reactance that account for posi-
tive as well as negative effects of personalization on prosocial outcomes (favourable
attitude change and helping intentions toward stigmatized social groups). Particular
promise lies in the finding that self-referencing served as a gateway to empathy and
prosocial outcomes. If personalized charity advertising is successful in prompting self-
reference, it can reinforce an empathic and altruistic frame of mind that makes recipi-
ents care about the well-being of others in need. However, it seems that in order to
fully harness the potential of personalized charity advertising, issues of reactance need
to be taken into account, and careful pretesting of campaigns is recommendable to
avoid ‘getting too personal’ – as White et al. (2008) have put it.
Disclosure statement
No potential conflict of interest was reported by the author(s).
References
Andrews, J.C., S. Durvasula, and S.H. Akhter. 1990. A framework for conceptualizing and measur-
ing the involvement construct in advertising research. Journal of Advertising 19, no. 4: 27–40.
http://www.jstor.org/stable/4188777.
Bagozzi, R.P., and D.J. Moore. 1994. Public service advertisements: Emotions and empathy guide
prosocial behavior. Journal of Marketing 58, no. 1: 56–70. doi:10.2307/1252251.
Bartsch, A., and F.M. Schneider. 2014. Entertainment and politics revisited: How non-escapist
forms of entertainment can stimulate political interest and information seeking. Journal of
Communication 64, no. 3: 369–96. doi:10.1111/jcom.12095.
Basil, D.Z., N.M. Ridgway, and M.D. Basil. 2006. Guilt appeals: The mediating effect of responsibil-
ity. Psychology & Marketing 23, no. 12: 1035–54. doi:10.1002/mar.20145.
Batson, C.D. 1987. Prosocial motivation: Is it ever truly altruistic? Advances in Experimental Social
Psychology 20: 65–122. doi:10.1016/S0065-2601(08)60412-8.
Batson, C.D., J.G. Batson, J.K. Slingsby, K.L. Harrell, H.M. Peekna, and R.M. Todd. 1991. Empathic
joy and the empathy-altruism hypothesis. Journal of Personality and Social Psychology 61, no.
3: 413–26. doi:10.1037/0022-3514.61.3.413.
Batson, C.D., J. Chang, R. Orr, and J. Rowland. 2002. Empathy, attitudes, and action: Can feeling
for a member of a stigmatized group motivate one to help the group? Personality and Social
Psychology Bulletin 28, no. 12: 1656–66. doi:10.1177/014616702237647.
Batson, C.D., M.P. Polycarpou, E. Harmon-Jones, H.J. Imhoff, E.C. Mitchener, L.L. Bednar, T.R.
Klein, and L. Highberger. 1997. Empathy and attitudes: Can feeling for a member of a stigma-
tized group improve feelings toward the group? Journal of Personality and Social Psychology
72, no. 1: 105–18. doi:10.1037/0022-3514.72.1.105.
INTERNATIONAL JOURNAL OF ADVERTISING 361
Batson, C.D., and A.A. Powell. 2003. Altruism and prosocial behavior. Vol. 5 Handbook of psych-
ology: Personality and social psychology, ed. Theodore Millon and Melvin Lerner, 463–84.
Hoboken, NJ: Wiley.
Bennett, R. 1997. ‘They shouldn’t let them out for us to see’: Empathy and affect intensity as
determinants of responses to representations of the facially disfigured in charity advertising.
International Journal of Nonprofit and Voluntary Sector Marketing 2, no. 3: 216–32. doi:10.
1002/nvsm.6090020304.
Bennett, R. 2015. Individual characteristics and the arousal of mixed emotions: Consequences for
the effectiveness of charity fundraising advertisements. International Journal of Nonprofit and
Voluntary Sector Marketing 20, no. 2: 188–209. doi:10.1002/nvsm.1500.
Brehm, J.W. 1966. A theory of psychological reactance. New York: Academic Press.
Brunel, F.F., and M.R. Nelson. 2000. Explaining gendered responses to “help-self” and “help-oth-
ers” charity ad appeals: The mediating role of world-views. Journal of Advertising 29, no. 3:
15–28. http://www.jstor.org/stable/4189149.
Burgess, D., B. Haney, M. Snyder, J.L. Sullivan, and J.E. Transue. 2000. Rocking the vote: Using
personalized messages to motivate voting among young adults. Public Opinion Quarterly 64,
no. 1: 29–52. doi:10.1086/316758.
Burnkrant, R.E., and H.R. Unnava. 1989. Self-referencing: A strategy for increasing processing of
message content. Personality and Social Psychology Bulletin 15, no. 4: 628–38. doi:10.1177/
0146167289154015.
Burnkrant, R.E., and H.R. Unnava. 1995. Effects of self-referencing on persuasion. Journal of
Consumer Research 22, no. 1: 17–26. doi:10.1086/209432.
Busselle, R., and H. Bilandzic. 2009. Measuring narrative engagement. Media Psychology 12, no. 4:
321–47. doi:10.1080/15213260903287259.
Chang, C. 2014. Guilt regulation: The relative effects of altruistic versus egoistic appeals for char-
ity advertising. Journal of Advertising 43, no. 3: 211–27. doi:10.1080/00913367.2013.853632.
Chang, C.-T. 2011. Guilt appeals in cause related marketing: The subversive roles of product
type and donation magnitude. International Journal of Advertising 30, no. 4: 587–616. doi:10.
2501/IJA-30-4-587-616.
Chang, C.-T., and Y.-K. Lee. 2010. Effects of message framing, vividness congruency and statis-
tical framing on responses to charity advertising. International Journal of Advertising 29, no. 2:
195–220. doi:10.2501/S0265048710201129.
Chang, C.-T., and Y.-K. Lee. 2011. The ‘I’ of the beholder: How gender differences and self-refer-
encing influence charity advertising. International Journal of Advertising 30, no. 3, 447–78.
doi:10.2501/IJA-30-3-447-478.
Cupchik, G.C. 1995. Emotion in aesthetics: Reactive and reflective models. Poetics 23: 177–88.
Das, E., P. Kerkhof, and J. Kuiper. 2008. Improving the effectiveness of fundraising messages: The
impact of charity goal attainment, message framing, and evidence on persuasion. Journal of
Applied Communication Research 36, no. 2: 161–75. doi:10.1080/00909880801922854.
Dillard, J.P., and L. Shen. 2005. On the nature of reactance and its role in persuasive health com-
munication. Communication Monographs 72, no. 2: 144–68. doi:10.1080/03637750500111815.
Eisenberg, N., R.A. Fabes, M. Schaller, P. Miller, G. Carlo, R. Poulin, C. Shea, and R. Shell. 1991.
Personality and socialization correlates of vicarious emotional responding. Journal of
Personality and Social Psychology 61, no. 3: 459–70. doi:10.1037/0022-3514.61.3.459.
Eisenberg, N., M. Wentzel, and J.D. Harris. 1998. The role of emotionality and regulation in
empathy-related responding. School Psychology Review 27, 506–22.
Escalas, J.E., M.C. Moore, and J.E. Britton. 2004. Fishing for feelings? Hooking viewers helps!
Journal of Consumer Psychology 14, no. 1–2: 105–14. doi:10.1207/s15327663jcp1401&2_12.
Fleming, M.A., R.E. Petty, and P.H. White. 2005. Stigmatized targets and evaluation: Prejudice as
a determinant of attribute scrutiny and polarization. Personality and Social Psychology Bulletin
31, no. 4: 496–507. doi:10.1177/0146167204271585.
Garcia, S.M., K. Weaver, G.B. Moskowitz, and J.M. Darley. 2002. Crowded minds: The implicit
bystander effect. Journal of Personality and Social Psychology 83, no. 4: 843–53. doi:10.1037//
0022-3514.83.4.843.
362 A. BARTSCH AND A. KLOß
Green, M.C., and T.C. Brock. 2000. The role of transportation in the persuasiveness of public nar-
ratives. Journal of Personality and Social Psychology 79, no. 5: 701–21. doi:10.1037/0022-3514.
79.5.701.
Griffin, M., B.J. Babin, J.S. Attaway, and W.R. Darden. 1993. Hey you, can ya spare some change?
The case of empathy and personal distress as reactions to charitable appeals. Advances
in Consumer Research 20, no. 1: 508–14. http://acrwebsite.org/volumes/7506/volumes/v20/
NA-20.
Hastall, M., and F. Sukalla. 2012. Der Einfluss verschiedener Arten von Botschaftsframing und dis-
positioneller Sensitivit€at fu
€r appetitive und aversive Reize der Rezipienten auf die Wirksamkeit
von Gesundheitsplakaten. Paper presented at the 6th conference media and health communi-
cation network, Munich.
Hayes, A.F. 2009. Beyond Baron and Kenny: Statistical mediation analysis in the new millennium.
Communication Monographs 76, no. 4: 408–20. doi:10.1080/03637750903310360.
Hong, S.-M., and S. Faedda. 1996. Refinement of the Hong psychological reactance
scale. Educational and Psychological Measurement 56, no. 1: 173–82. doi:10.1177/
0013164496056001014.
Johnson, B.T., and A.H. Eagly. 1989. Effects of involvement on persuasion: A meta-analysis.
Psychological Bulletin 106, no. 2: 290–314. doi:10.1037/0033-2909.106.2.290.
Kalyanaraman, S., and S.S. Sundar. 2006. The psychological appeal of personalized content in
web portals: Does customization affect attitudes and behavior? Journal of Communication 56,
no. 1: 110–32. doi:10.1111/j.1460-2466.2006.00006.x.
Kim, N. 2014. Advertising strategies for charities. International Journal of Advertising 33, no. 4:
707–24. doi:10.2501/IJA-33-4-707-724.
Kreuter, M.W., F.C. Bull, E.M. Clark, and D.L. Oswald. 1999. Understanding how people process
health information: A comparison of tailored and nontailored weight-loss materials. Health
Psychology 18, no. 5: 487–94. doi:10.1037/0278-6133.18.5.487.
Latane, B., and S. Nida. 1981. Ten years of research on group size and helping. Psychological
Bulletin 89, no. 2: 308–24. doi:10.1037/0033-2909.89.2.308.
Lee, S., K.P. Winterich, and W.T. Ross Jr. 2014. I'm moral, but I won’t help you: The distinct roles
of empathy and justice in donations. Journal of Consumer Research 41: 678–96. doi:10.1086/
677226.
Lester, H., and H. Pattison. 2000. Development and validation of the attitudes towards the
homeless questionnaire. Medical Education 34, no. 4: 266–68. doi:10.1046/j.1365-2923.2000.
00495.x.
Mares, M.-L., and E. Woodard. 2005. Positive effects of television on children’s social interactions:
A meta-analysis. Media Psychology 7, no. 3: 301–22. doi:10.1207/S1532785XMEP0703_4.
Maslowska, E.H. 2011. Does Dear John always work? Personalized communication revisited.
Paper presented at the annual meeting of the International Communication Association (ICA),
TBA, Boston, MA.
Masthoff, J., S. Langrial, and K. van Deemter. 2013. Personalizing triggers for charity actions.
Persuasive technology, ed. Shlomo Berkovsky and Jill Freyne, 125–36. Berlin, Heidelberg:
Springer. doi:10.1007/978-3-642-37157-8_16.
Mehrabian, A., and N. Epstein. 1972. A measure of emotional empathy. Journal of Personality 40,
no. 4: 525–43. doi:10.1111/j.1467-6494.1972.tb00078.x.
Nuttin, J.M. 1985. Narcissism beyond gestalt and awareness: The name letter effect. European
Journal of Social Psychology 15, no. 3: 353–61. doi:10.1002/ejsp.2420150309.
Oliver, M.B., J.P. Dillard, K.B. Keunmin, and D.J. Tamul. 2012. The effect of narrative news format
on empathy for stigmatized groups. Journalism & Mass Communication Quarterly 89, no. 2:
205–24. doi:10.1177/1077699012439020.
Peng, W., M. Lee, and C. Heeter. 2010. The effects of a serious game on role-taking and willing-
ness to help. Journal of Communication 60, no. 4: 723–42. doi:10.1111/j.1460-2466.2010.
01511.x.
Petrescu, M., and P. Korgaonkar. 2011. Viral advertising: Definitional review and synthesis.
Journal of Internet Commerce 10, no. 3: 208–26. doi:10.1080/15332861.2011.596007.
INTERNATIONAL JOURNAL OF ADVERTISING 363
Petty, R.E., J. Barden, and S.C. Wheeler. 2002. The elaboration likelihood model of persuasion:
Health promotions that yield sustained behavioral change. Emerging theories in health promo-
tion practice and research, ed. Ralph J. DiClemente, Richard A. Crosby, and Michelle Kegler,
71–99. San Fransisco: Jossey-Bass.
Petty, R.E., and J.T. Cacioppo. 1986. Communication and persuasion: Central and peripheral routes
to attitude change. New York: Springer.
Petty, R.E., J.T. Cacioppo, and D. Schumann. 1983. Central and peripheral routes to advertising
effectiveness: The moderating role of involvement. Journal of Consumer Research 10: 135–46.
Petty, R.E., S.C. Wheeler, and G.Y. Bizer. 2000. Attitude functions and persuasion: An elaboration
likelihood approach to matched versus mismatched messages. Why we evaluate: Functions of
attitudes, ed. Gregory R. Maio and James M. Olson, 133–62. Mahwah, NJ: Erlbaum.
Rains, S.A. 2013. The nature of psychological reactance revisited: A meta-analytic review. Human
Communication Research 39, no. 1: 47–73. doi:10.1111/j.1468-2958.2012.01443.x.
Ryan, R.M., and E.L. Deci. 2000. Self-determination theory and the facilitation of intrinsic motiv-
ation, social development, and well-being. The American Psychologist 55, no. 1: 68–78. doi:10.
1037/0003-066X.55.1.68.
Shen, L. 2011. The effectiveness of empathy- versus fear-arousing antismoking PSAs. Health
Communication 26, no. 5: 404–15. doi:10.1080/10410236.2011.552480.
Slater, M.D., and D. Rouner. 2002. Entertainment-education and elaboration likelihood:
Understanding the processing of narrative persuasion. Communication Theory 12, no. 2:
173–91. doi:10.1111/j.1468-2885.2002.tb00265.x.
Small, D.A., and N.M. Verrochi. 2009. The face of need: Facial emotion expression on charity
advertisements. Journal of Marketing Research 46, no. 6: 777–87. doi:10.1509/jmkr.46.6.777.
Tukachinsky, R., and R. Stokunaga. 2013. The effects of engagement with entertainment. Annals
of the International Communication Association 37, no. 1: 287–322. doi:10.1080/23808985.
2013.11679153.
Vesanen, J. 2007. What is personalization? A conceptual framework. European Journal of
Marketing 41, no. 5/6: 409–18. doi:10.1108/03090560710737534.
Weinstein, N., and R.M. Ryan. 2010. When helping helps: Autonomous motivation for prosocial
behavior and its influence on well-being for the helper and recipient. Journal of Personality
and Social Psychology 98, no. 2: 222–44. doi:10.1037/a0016984.
White, T.B., D. Zahay, H. Thorbjørnsen, and S. Shavitt. 2008. Getting too personal: Reactance to
highly personalized email solicitations. Marketing Letters 19, no. 1: 39–50. doi:10.1007/s11002-
007-9027-9.
Wirth, W., M. Hofer, and H. Schramm. 2012. Beyond pleasure: Exploring the eudaimonic enter-
tainment experience. Human Communication Research 38, no. 4: 406–28. doi:10.1111/j.1468-
2958.2012.01434.x.
Zaichkowsky, J.L. 1986. Conceptualizing involvement. Journal of Advertising 15, no. 2: 4–34. doi:
10.1080/00913367.1986.10672999.
Zillmann, D., and J. Cantor. 1977. Affective responses to the emotions of a protagonist. Journal
of Experimental Social Psychology 13, no. 2: 155–65. doi:10.1016/S0022-1031(77)80008-5.