Research article
Women’s reactions to ingroup members who protest discriminatory
treatment: The importance of beliefs about inequality and response
appropriateness
DONNA M. GARCIA1*, MICHAEL T. SCHMITT2
,
NYLA R. BRANSCOMBE3
AND NAOMI ELLEMERS4
1
University of Guelph, Canada
2
Simon Fraser University, Canada
3
University of Kansas, USA
4
Leiden University, The Netherlands
Abstract
Our goal was to identify factors that shape women’s responses to ingroup members who protest gender discrimination. We
predicted and found that women who perceived gender discrimination as pervasive regarded a protest response as being
more appropriate than a no protest response and expressed greater liking and less anger towards a female lawyer who
protested rather than did not protest an unfair promotion decision. Further, beliefs about the appropriateness of the
response to discrimination contributed to evaluations of the protesting lawyer. Perceptions that the complaint was an
appropriate response to the promotion decision led to more positive evaluations of an ingroup discrimination protester.
Copyright # 2009 John Wiley & Sons, Ltd.
Protest can be an effective means of improving the plight of a devalued group. Historically, there are many examples of
protest, even from a single individual, that have advanced a group’s social position (e.g. Meritor Savings Bank vs. Vinson,
477 US 57, 1986; Dekker vs. VJV-Centrum ECJ, 1992). Despite the potential gains to be obtained by protesting
illegitimate treatment, protestors might not always be appreciated by members of their own group. Whether disadvantaged
group members respond positively or negatively to ingroup protestors will likely depend upon the perceived implications
that the protestor’s action has for the ingroup. Unless protest is seen as justiﬁed by the social circumstances and an effective
means of bringing about positive change, a protestor might be seen as making the ingroup look like complainers. Such
threat to the ingroup’s reputation could evoke the ire and disdain of the disadvantaged group towards the protestor. Hence,
perceptions of the justiﬁcation for and likely consequences of protest will be critical to others’ reactions to an ingroup
discrimination claimer. We propose that protest by an ingroup member will be seen as appropriate and thus appreciated to
the extent that observers perceive that their ingroup is targeted by pervasive discrimination.
SOCIAL CONSEQUENCES OF CLAIMING DISCRIMINATION
Gender discrimination continues to be widespread throughout Western employment settings (see Charles & Grusky,
2004). The continuation of gender discrimination has substantive negative implications for women’s economic and
European Journal of Social Psychology
Eur. J. Soc. Psychol. 40, 733–745 (2010)
Published online 6 July 2009 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/ejsp.644
*Correspondence to: Donna M. Garcia, Department of Psychology, University of Guelph Ontario, Guelph, Ontario, N1G 2W1Canada.
E-mail: donnagarcia3@gmail.com
Copyright # 2009 John Wiley & Sons, Ltd.
Received 19 October 2008
Accepted 7 April 2009
psychological outcomes (Major & O’Brien, 2005; Schmitt & Branscombe, 2002). Moreover, concern about
discrimination and social devaluation in male-dominated domains can have a demoralizing effect on women’s
aspirations (Gupta & Bhawe, 2007), comfort (Adams, Garcia, Purdie-Vaughns, & Steele, 2006) and performance (Steele,
Spencer, & Aronson, 2002). In sum, the evidence is overwhelmingly clear: gender discrimination is detrimental to women.
Despite the many costs of discrimination, members of disadvantaged groups, including women, often do not publicly
complain about unfair treatment (Dixon, Storen, & Horn, 2002; Stangor, Swim, Van Allen, & Sechrist, 2002; Swim &
Hyers, 1999). In survey (Dixon et al., 2002) and interview (Feagin & Sikes, 1994) research, respondents indicate that one
key reason they keep silent about experiences with discrimination is that they fear they will be disliked or labelled a whiner
and troublemaker. Laboratory studies show that low status group members tend to be reluctant to acknowledge an
experience with discrimination in the presence of others (Postmes, Branscombe, Spears, & Young, 1999; Stangor et al.,
2002; Swim & Hyers, 1999). These ﬁndings provide evidence that low status group members fear claiming discrimination
because of the potential social costs; but, are such fears warranted?
Concerns about the social costs of claiming discrimination are valid. In laboratory studies, people tend to derogate
individuals who publicly label an experience as discrimination (Garcia, Horstman Reser, Amo, Redersdorff, &
Branscombe, 2005; Kaiser & Miller, 2001). Much of this ‘social costs’ research has focussed on people’s reactions to
outgroup discrimination claimers (e.g. men’s derogation of a woman who claims discrimination). Less research attention
has been paid to evaluations of discrimination claims by ingroup members, despite evidence that people are inﬂuenced
more by ingroup rather than outgroup social pressure (Ellemers, Doosje, & Spears, 2004; Postmes & Branscombe, 2002).
Consequently, the present research focuses on responses to ingroup discrimination claimers. A second concern is that little
research has sought to identify the mechanisms underlying reactions to discrimination claimers, including the
circumstances that might lead to support versus derogation. Although some research has examined the role of justice
beliefs (Kaiser, Dyrenforth, & Hagiwara, 2006), it has focussed on how this factor affects reactions to outgroup
discrimination claimers. Thus, we examine two factors that social identity theory (Tajfel & Turner, 1986) predicts will
inﬂuence responses to ingroup discrimination claimers: beliefs about pervasiveness of discrimination and the perceived
appropriateness of the complaint in the speciﬁc situation.
Ingroup evaluations
Judgments of ingroup members tend to be more extreme than those of outgroup members (Branscombe, Wann, Noel, &
Coleman, 1993; Marques, Yzerbyt, & Leyens, 1988). Relative to members of the outgroup, people are more likely to
praise likable and norm-upholding ingroup members and derogate dislikeable and norm-violating ingroup members. By
lauding ‘good’ ingroup members and castigating ingroup ‘black sheep,’ people can afﬁrm the high standards and
‘normative legitimacy’ of their ingroup and maintain a positive ingroup identity. The proclivity for people to respond most
extremely to ingroup members suggests that they may be particularly vigilant in monitoring ingroup responses to
discrimination. Whether they support or derogate an ingroup discrimination claimer should depend on whether they
perceive such protest as beneﬁting or harming their ingroup.
The few studies that have examined responses to ingroup discrimination claimers have produced mixed evidence. Some
research has shown that people are even harsher in their evaluation of ingroup discrimination claimers (Garcia et al.,
2005). In contrast, Dodd, Giuliano, Boutell, and Moran (2001) observed ingroup social support for a woman who
confronted rather than ignored a man who made a sexist comment. However, women only liked and respected the
confronter more than a non-confronter when the comment was blatantly sexist. When the remark was ambiguous,
women’s evaluations were unaffected by the target’s response to the remark.
The moderating role of ambiguity found by Dodd et al. (2001) provides insight into the discrepancy between the
ingroup social support found in their study and the ingroup social costs found by other researchers. In the research by
Garcia et al. (2005), participants read about an individual in a prior research study who had failed a creativity test.
Although participants learned that the test-evaluator in that study was biased against the target’s gender group, it is not
certain that they attributed the failure to discrimination: a biased evaluator might not always lead to a biased evaluation,
especially on standardized tests that could be assumed to use uniform grading criteria. Garcia et al. conjectured that if the
discrimination had been less ambiguous (as in the Dodd et al. blatant condition), their results might have shown support for
an ingroup discrimination claimer. This possibility highlights the role that perceptions of discrimination might have for
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
734 Donna M. Garcia et al.
observers’ reactions to ingroup discrimination claimers. Although the research by Dodd et al. relates to perceptions of
discrimination concerning a speciﬁc event, we propose that perceptions of discrimination in general should also play an
important role in evaluations of ingroup discrimination claimers.
Perceptions of Pervasive Discrimination against Women
The social identity approach emphasizes that women’s subjective understanding of gender inequality will shape their
beliefs about the appropriate response to structural differences between men and women (Tajfel, 1978). Accordingly,
perceiving that their gender group experiences unfair devaluation will encourage women to self-deﬁne in terms of their
gender group membership and support strategies that promote women’s collective advancement and challenge the status
quo (Ellemers, 1993). In contrast, women who feel that gender discrimination is uncommon and that status boundaries are
permeable will be inclined to think in terms of their individual self and endorse individual mobility endeavours (Wright,
Taylor, & Moghaddam, 1990). These predictions about the importance of structural beliefs for collective mobility among
disadvantaged group members have been well-supported (see Branscombe & Ellemers, 1998; Ellemers, 1993; Simon,
Hastedt, & Aufderheide, 1997). We argue that variations in subjective understanding of gender relations should also
inﬂuence beliefs regarding the appropriate response to individual experiences with discrimination.
Research regarding the psychological consequences of perceiving discrimination against one’s group has focussed on
how these perceptions affect individual well being (see Major & O’Brien, 2005; Schmitt & Branscombe, 2002) or
collective action tendencies (Wright et al., 1990). The present investigation considers how perceptions of discrimination
inﬂuence observers’ reactions to ingroup discrimination claimers. Like collective action, protest from individuals
challenges unfair treatment; thus, we reason that beliefs about the pervasiveness of gender inequality will also inﬂuence
support for a woman who protests discrimination. Following from social identity theory, we propose that support for an
ingroup member who protests discrimination should be greatest among women who perceive gender discrimination as
pervasive. Further, the extent to which perceptions of pervasiveness lead to support is likely to be a function of the
perceived appropriateness of the protest in the given circumstances.
Appropriateness of a Response to Discrimination
Drawing on Kowalski’s (1996) analysis of complaining, Garcia et al. (2005) conjectured that evaluations of ingroup
members who claim discrimination would depend on the perceived appropriateness of the complaint. Although people
often dislike complainers, they often support a complainer if the complaint is seen as appropriate (e.g. warranted,
legitimate, impactful). Likewise, the more legitimate people think that collection action is as a means to foster positive
change, the more inclined they are to support the action (Hornsey et al., 2006; Wright et al., 1990). Thus, people’s
responses to ingroup discrimination claimers should be shaped by the perceived appropriateness of the complaint. On this
basis, we anticipate that women who believe sexism is pervasive are likely to regard protest as an appropriate way for
group members to deal with unfair treatment, which in turn will lead them to favour an ingroup protestor.
Type of protest
Just as women can choose either individual or collective mobility strategies in response to gender inequality (Ellemers,
1993), they can respond to a personal experience with discrimination by expressing either individual or collective justice
concerns. A collective approach to protest is one in which a woman responds to discriminatory treatment by pointing out
the harm and unfairness to women in general and emphasizing the assets and capabilities of women as a whole. An
individual approach to protest is one in which a woman responds by focussing on the harm and unfairness to herself and
emphasizing her personal deservingness. We included both forms of protest to assess whether the type of protest matters.
Although it is possible that women who perceive sexist discrimination as pervasive will favour an ingroup member who
promotes women’s collective advancement, it is also feasible that they will respond favourably to either form of protest.
Regardless of the type of protest, an individual woman who confronts unfair treatment could be regarded as inspiring other
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
Women’s reactions to ingroup protestors 735
women to do the same and opening the door for future women by overcoming the barriers in her way. Hence, women who
perceive their gender group as disadvantaged could endorse both forms of protest to the extent they believe that protest is
more appropriate than no protest in the given context.
Research Hypotheses
The current research addressed three questions. First, we examined whether women’s emotional and evaluative reactions
to a female target who either protested or did not protest a biased outcome would depend upon beliefs about the
pervasiveness of gender discrimination. We predicted that evaluations of both a protesting target and her response would
be positive among those women who regard discrimination as pervasive, but negative among those who perceive
discrimination as rare. Second, we tested whether the moderating effects of pervasiveness of discrimination on reactions to
a protesting target would be mediated by beliefs about the appropriateness of the response. We predicted that women who
perceive gender discrimination as pervasive would respond positively to a woman who protested an unfair hiring decision
because they would regard a protest response as being more appropriate than a non-protest response. Third, we examined
whether evaluations of a protesting target and her action would vary depending upon the form of protest she took—
individual or collective. We examined whether women would favour either (a) an ingroup protestor who focussed on
collective justice rather than individual justice concerns or (b) any ingroup protestor because both forms of protest would
be better than no protest in the given circumstances.
To examine these questions, we had women read about an unfair hiring decision in which a law ﬁrm chose not to grant a
highly qualiﬁed female lawyer partnership in the ﬁrm and instead promoted a less qualiﬁed male lawyer. We varied
whether the female lawyer chose to protest the decision or not. We also considered two different types of protest—an
individual strategy in which she focussed on the unfairness of the hiring decision for herself or a collective strategy where
she emphasized the ﬁrm’s unfairness to women. After reading the scenario, participants indicated their anger and liking for
this female lawyer, and their perceptions concerning the appropriateness of her response to the hiring decision.
METHOD
Participants and Cover Story
We recruited 129 female undergraduates at the University of Kansas to participate in a study concerning ‘impressions
that people form of others in various life situations.’ Participants received credit towards their introductory psychology
course.
Materials and Procedure
Participants received a booklet containing the description of an event that ostensibly occurred in an American law ﬁrm and
a series of questions concerning that event. All participants learned about an incident in which a male lawyer was
promoted to senior partner over a more qualiﬁed female lawyer. The decision was clearly discriminatory. The female
lawyer won more high proﬁle cases, won greater average awards for her clients, brought more money into the ﬁrm and
worked more hours than her male colleague. His assets were that he played golf well and some of the senior partners
enjoyed the company of the male lawyer more and did not think women were cut out to be lawyers.
After reading that the female lawyer felt the decision was unfair, participants read one of three versions regarding her
response to the discriminatory promotion decision. They were randomly assigned to the No Protest (N ¼ 41), Individual
Protest (N ¼ 43) or Collective Protest (N ¼ 45). In the No Protest condition, the female lawyer said that, despite her
disappointment, she would learn to live with the ﬁrm’s decision. She stated,
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
736 Donna M. Garcia et al.
The decision has been made and there’s little that I can do about it at this point. Because I can’t change things, I’m going
to keep on working at the ﬁrm and make the best of my current position.
In the Individual Protest condition, the target decided to ask the ﬁrm to reconsider the decision and said
This ﬁrm did not look at my individual qualities sufﬁciently. I agree that most women are not cut out for partnership, but
I am different from other women. I expect only to be judged on my individual qualiﬁcations. This ﬁrm is treating me
unfairly.
In the Collective Protest condition, the target also decided to ask the ﬁrm to reconsider their decision and said
This ﬁrm is not fully considering the abilities of women. If they did, they would see that women could perform just as
competently as men. I expect women to be treated equally to men. This ﬁrm is treating women unfairly.
Pervasiveness of Gender Discrimination
Participants completed the Modern Sexism Scale to assess their beliefs about the pervasiveness of gender discrimination.
This 8-item scale includes statements such as ‘Discrimination against women is no longer a problem in the United States,’
‘Women often miss out on good jobs due to sexual discrimination,’ and ‘On average, people in our society treat men and
women equally.’ We averaged the items together, after the appropriate reverse-scoring, so that high scores indicated
perceptions of gender discrimination as pervasive (a ¼ .75).
Outcome Measures
Our outcome measures included participants’ beliefs about the promotion decision, emotional reactions to the female
lawyer, evaluations of her, and perceptions of the target’s response to the event. Participants responded to all measures
using Likert type scales, which unless speciﬁed otherwise ranged from 1 (Strongly disagree) to 7 (Strongly agree).
Anger Towards the Target
We assessed participants’ anger towards the female target with a single item: ‘I feel angry at Catherine.’ The higher the
score on this item, the greater amount of anger the participants expressed towards the target.
Evaluation of Target
Participants next evaluated the likeability of the target by completing six items. These items included: ‘I like Catherine,’ ‘I
admire Catherine,’ ‘Catherine is the type of person I would like to be friends with,’ ‘Catherine has many positive traits,’ ‘I
would like to be a coworker of Catherine,’ and ‘I feel proud of Catherine.’ We averaged these items together to create a
Liking measure, with high scores indicating greater liking (a ¼ .88).
Evaluation of Target’s Response
The next measure assessed evaluations of the target’s response to the partnership decision. Participants speciﬁed how
appropriate the target’s response was, how effective her response was for achieving her personal goals, how assertive her
response was, and how harmful her response was for other women (reverse-scored). We averaged these four items together
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
Women’s reactions to ingroup protestors 737
to create a Response Evaluation measure (a ¼ .76). High scores on this measure indicated that the participants regarded the
target’s response to the promotion decision as a positive way of dealing with the ﬁrm’s discrimination.
Unfairness of the Promotion Decision
We included three measures that assessed participants’ impressions about the unfairness of the partnership decision. First,
participants indicated who they thought was the more qualiﬁed candidate. Responses for this item were coded such that
high scores indicated the female lawyer was perceived as the more qualiﬁed candidate and low scores indicated that the
male lawyer was the more qualiﬁed candidate. Participants next completed a 9-item measure that assessed the extent to
which the decision to promote the male lawyer to partner was regarded as gender discrimination (a ¼ .87). Items on this
measure were: ‘Catherine was denied the promotion because of sex discrimination,’ ‘I do not think the event was a valid
instance of sex discrimination (reversed)’ and ‘I perceived this as a clear-cut case of discrimination.’ We averaged the
items together, with high scores indicating agreement that the promotion decision represented gender discrimination.
These latter two measures allowed us to test whether participants interpreted the scenario as we intended. If the mean
response on both measures is high, then participants correctly identiﬁed the female lawyer as the stronger candidate and
regarded the promotion decision as discrimination.
Manipulation Checks
We included two items to assess whether participants correctly understood Catherine’s response to the promotion decision.
Participants indicated the extent to which they agreed that ‘Catherine wanted to be seen as an individual’ and ‘Catherine
was concerned with how discrimination affects all women.’ Although we anticipated that responses to the items might be
higher to the items in both protest conditions compared to the no protest condition, we expected responses to the ﬁrst
question to be highest in the individual protest condition and responses to the second item to be highest in the collective
protest condition.
RESULTS
Manipulation and Event Interpretation Checks
To test whether participants understood the target response conditions as we intended, we analysed participants’
perceptions about why the female lawyer responded to the partnership decision as she did. Two one-way analyses of
variance (ANOVAs) revealed that participants’ perceptions of the female lawyer’s response to the partnership decision
varied across conditions for both the concern with individuality, F(2, 126) ¼ 11.90, p < .001, h2
¼ .29, and the concern for
women, F(2, 126) ¼ 14.43, p < .0001, h2
¼ .32, manipulation checks. Participants in the individual protest condition were
more likely to agree that the female target wanted to be seen as an individual (M ¼ 6.49, SD ¼ .77) than were participants
in the collective protest (M ¼ 5.42, SD ¼ 1.50), t(126) ¼ 3.68, p < .001, h2
¼ .31 or the no protest (M ¼ 5.12, SD ¼ 1.65)
conditions, t(126) ¼ À4.61, p < .001, h2
¼ .38. In contrast, participants in the collective protest condition (M ¼ 4.84,
SD ¼ 1.67) regarded the target as more concerned with how discrimination affects all women than those in the individual
(M ¼ 3.77, SD ¼ 1.59), t(126) ¼ À3.07, p ¼ .003, h2
¼ .26 or the no protest conditions (M ¼ 2.95, SD ¼ 1.67),
t(126) ¼ À5.34, p < .001, h2
¼ .43. Participants in the individual condition also perceived the target as more concerned
with the effects of discrimination on women than did participants in the no protest condition, t(126) ¼ À2.28, p ¼ .024,
h2
¼ .20. This ﬁnding indicates that participants regarded both forms of protest as reﬂecting a concern with women as a
group to some extent.
We also examined whether participants’ interpretation of the information that they read about the partnership decision
corresponded with our intentions. That is, did participants perceive the female lawyer as the stronger candidate and did
they regard the decision to promote the male lawyer as discrimination? We found that participants accurately identiﬁed the
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
738 Donna M. Garcia et al.
female lawyer as the better candidate for the senior partnership position. The mean response (M ¼ 6.43, SD ¼ .88) to the
question ‘Who was more qualiﬁed to receive the partnership?’ (coded so that 1 ¼ Paul; 7 ¼ Catherine), was signiﬁcantly
greater than the midpoint of 4, t(128) ¼ 31.33, p < .001. Participants also perceived the ﬁrm’s decision as discrimination;
The mean response on the measure assessing whether the event reﬂected discrimination (M ¼ 5.29, SD ¼ 1.11) was
reliably greater than the midpoint of 4, t(128) ¼ 13.19, p < .001. Judgments concerning who was the best candidate, F(2,
126) ¼ .07, p > .25, and whether the decision was discriminatory, F(2, 126) ¼ .16, p > .25, did not differ by response
condition.1
Regression Tests for Moderation
The main goals of the moderated multiple regressions were to examine (a) whether the target’s response to the hiring
decision inﬂuenced evaluations of the target and her action and (2) whether these effects were moderated by beliefs about
the pervasiveness of gender discrimination.2
Thus, we subjected all dependent measures to separate moderated multiple
regressions (MMR) with response type, pervasiveness of sexism beliefs, and their interaction simultaneously entered as
predictor variables. Following Cohen, Cohen, West, and Aiken (2003), we contrast coded the three-level protest type
variable to represent the two contrasts of interest. ‘Response Choice’ contrasted the no protest group versus the two protest
groups (no protest ¼ À2/3, individual protest ¼ 1/3 and collective protest ¼ 1/3) and ‘Protest Type’ contrasted the
collective protest group versus the individual protest group (no protest ¼ 0, individual protest ¼ À1/2 and collective
protest ¼ 1/2). Relative to dummy or effect coding, contrast coding is more appropriate when testing speciﬁc hypotheses
because it allows one to compare orthogonal combinations of a categorical variable and provides greater statistical power
(Cohen et al., 2003). After centring the continuous pervasiveness of sexism moderator variable, we computed the cross
products of Response Choice and Pervasiveness of Sexism and Protest Type and Pervasiveness of Sexism to produce the
two interaction terms required to represent the interaction between the experimental condition and the moderator variable.
We then regressed each outcome variable simultaneously onto the two contrast variables, the continuous Pervasiveness of
Sexism measure, and the two interaction terms (Cohen et al., 2003).3
Anger Towards the Target
Following the procedures outlined above, we regressed anger towards the target on the contrast and moderator variables.
Response Choice reliably predicted anger, B ¼ À1.43, t(123) ¼ À5.02, p < .001: Participants expressed greater anger
towards a target who did not protest (M ¼ 3.07; SD ¼ 2.07) than one who did protest (M ¼ 1.68; SD ¼ 1.27). Of greater
interest, was the signiﬁcant interaction between Response Choice and Pervasiveness of Sexism, B ¼ À1.08,
t(123) ¼ À2.92, p ¼ .004. As predicted, whether anger was greater for a non-protesting or protesting target depended
upon participants’ beliefs about the pervasiveness of discrimination. To understand the nature of this interaction, we
computed simple slopes. As illustrated in Figure 1, the more sexism was perceived as pervasive, the greater anger was
towards a non-protesting target, B ¼ .74, t(123) ¼ 2.40, p ¼ .018, and less anger towards a protesting target, B ¼ À.35,
t(123) ¼ À1.82, p ¼ .072, though the latter effect was only marginally signiﬁcant. Response Choice was the only contrast
that produced signiﬁcant effects. The type of protest that the target chose produced neither a main effect, B ¼ À.18,
t(123) ¼ À.54, p ¼ .589, nor an interaction with beliefs about pervasiveness of discrimination, B ¼ .18, t(123) ¼ .44,
p ¼ .662.
1
Two separate moderated multiple regression analyses (following the same procedure used to analyse the main dependent measures) produced neither a
reliable main effect of response condition nor a signiﬁcant interaction between response condition and the pervasiveness of sexism measure ps ! .20. The
moderated multiple regression analyses produced one signiﬁcant ﬁnding for each of the measures: beliefs about the pervasiveness of sexism were
positively related to participants’ beliefs that the female lawyer was the more qualiﬁed candidate b ¼ .41, t(123) ¼ 4.96, p < .001 and their perceptions
that the promotion outcome was discriminatory, b ¼ .53, t(123) ¼ 6.94, p < .001.
2
Before testing for the moderating effects of perceived pervasiveness of sexism, we ran a one-way ANOVAwith response type as the independent variable
and pervasiveness of sexism as the dependent variable. This analysis was not signiﬁcant, F(2, 125) ¼ .782, p ¼ .460, indicating that pervasiveness of
sexism beliefs were unaffected by the response type manipulation.
3
We report unstandardized coefﬁcients (B) rather that betas (b). Because bs are not properly standardized in interaction terms, they are not interpretable
whereas B represents the difference between the unweighted means of the groups involved in the contrast (see Cohen et al., 2003).
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
Women’s reactions to ingroup protestors 739
Evaluation of Target
We followed the same MMR procedure as we did with the anger measure, this time regressing Liking onto the contrast and
moderator variables. This analysis revealed a main effect of Response Choice, B ¼ .49, t(123) ¼ 2.57, p ¼ .011, such that
liking was greatest for the target when she chose to protest (M ¼ 5.79; SD ¼ 0.88) rather than not protest the hiring
decision (M ¼ 5.31; SD ¼ 1.30). This main effect was again qualiﬁed by a signiﬁcant interaction between Response
Choice and Pervasiveness of Sexism, B ¼ .84, t(123) ¼ 3.40, p ¼ .001. As illustrated in Figure 2, greater pervasiveness of
sexism was associated with less liking for the target when she did not protest the promotion decision, B ¼ À.47,
t(123) ¼ À2.30, p ¼ .023, and greater liking for the target when she protested the promotion decision, B ¼ .36,
t(123) ¼ 2.67, p ¼ .009. Consistent with the anger measure, no effects of Protest Type emerged from these analyses either
on its own, B ¼ .01, t(123) ¼ À.025, p ¼ .980, or in combination with the pervasiveness of discrimination measure,
B ¼ .12, t(123) ¼ .46, p ¼ .650.
Figure 1. Beliefs about the pervasiveness of sexism moderate the relationship between the target’s response to the discrimination and
anger towards the target. Ã
p < .05; y p ¼ .07
Figure 2. Beliefs about the pervasiveness of sexism moderate the relationship between the target’s response to the discrimination and
liking for the target. Ã
p < .05
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
740 Donna M. Garcia et al.
Evaluation of Target’s Response
We next submitted the response appropriateness measure to the MMR analyses. Once again, we found a main effect of
Protest Choice, B ¼ 1.44, t(123) ¼ 6.66, p < .001: Participants perceived a protest response as being more appropriate
(M ¼ 5.31; SD ¼ 1.30) than a no protest response (M ¼ 3.88; SD ¼ 1.46). This main effect was qualiﬁed by a signiﬁcant
interaction of Response Choice and Pervasiveness of Sexism, B ¼ .86, t(123) ¼ 3.04, p ¼ .003. As beliefs about the
pervasiveness of sexism increased, a no protest response was rated less appropriate, B ¼ À.53, t(123) ¼ À2.25, p ¼ .026,
whereas a protest response was rated as more appropriate B ¼ .32, t(123) ¼ 2.04, p ¼ .043 (see Figure 3). The Protest Type
contrast produced one marginal effect. Participants tended to rate an individual protest as being more appropriate than a
collective strategy, B ¼ À.41, t(123) ¼ À1.73, p ¼ .087. However, the Protest Type X Pervasiveness of Discrimination
interaction was not signiﬁcant, B ¼ .25, t(123) ¼ .79, p ¼ .433.
Tests of Mediated Moderation
Our next goal was to test our mediated moderation model. Mediated moderation occurs when two predictor variables (in
our case: Response Type and Pervasiveness of Discrimination) interactively affect a mediator, which in turn inﬂuences an
outcome variable (Morgan-Lopez & MacKinnon, 2006). The previous MMR analyses established that the Response
Choice Â Pervasiveness of Discrimination interaction was associated with Response Appropriateness (mediator) and both
Anger and Liking. A regression test conﬁrmed that the mediator was related to both Anger, b ¼ À.53, t(127) ¼ À7.11,
p < .001, and Liking, b ¼ .49, t(127) ¼ 6.41, p < .001. When Response Appropriateness was added to the MMR models,
the Response Choice Â Pervasiveness of Discrimination interaction appeared to reduce for both Anger, B ¼ À.63,
t(122) ¼ À1.78, p ¼ .077 and Liking, B ¼ .52, t(122) ¼ 2.25, p ¼ .026, but Evaluation Response remained unchanged and
signiﬁcant for both, B ¼ À.53, t(122) ¼ À4.81, p < .001 and B ¼ .37, t(122) ¼ 5.10, p < .001, respectively. To test this
polential mediation effect, we followed the bootstrapping method (with 1000 iterations) advocated by Preacher, Rucker
and Hayes (2007; see also Muller, Judd, & Yzerbyt, 2005).4
This procedure tests the null hypothesis that the indirect path
from the interaction term to the dependent variable via the mediator does not signiﬁcantly differ from zero. If zero is not
Figure 3. Beliefs about the pervasiveness of sexism moderate the relationship between the target’s response to the discrimination and
the perceived response appropriateness. Ã
p < .05
4
We chose bootstrapping methodology as a means of testing mediation over the commonly used Sobel test. Because the Sobel test assumes large sample
sizes and that the indirect effect is normally distributed, bootstrapping methodology was a more appropriate test of mediation (Muller et al., 2005;
Preacher et al., 2007).
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
Women’s reactions to ingroup protestors 741
contained within the conﬁdence intervals (CI) computed by the bootstrapping procedure, then one can conclude that the
indirect effect is indeed signiﬁcantly different from zero at p < 0.05. We adapted the SPSS macro provided by Preacher
and Hayes (2008) for tests of multiple mediation to test for mediated moderation. We entered the interaction between
Response Choice and Pervasiveness of Discrimination as the independent variable, Response Appropriateness as the
mediator and the main effects of Response Choice and Pervasiveness of Discrimination as covariates. Separate analyses
conﬁrmed the signiﬁcance of the indirect effect of the Response Type Â Pervasiveness of Discrimination interaction via
Response Appropriateness on both Anger (95% CI ¼ À0.9058 to À0.0830) and Liking (95% CI ¼ 0.0675 to 0.6776).
Because zero is not contained within either conﬁdence interval, the ﬁndings illustrate that the interactive effect of
Response Type and Pervasiveness of Discrimination on Anger and Liking for the target is a function of beliefs about the
appropriateness of the target’s response.
DISCUSSION
The present study examined two factors that could inﬂuence women’s reactions to an ingroup member who protested
unfair treatment: perceptions of pervasive gender discrimination and beliefs about the appropriateness of the protest.
Overall, we found that women expressed less anger and greater liking for a woman who protested an unfair hiring decision.
They also perceived protest as being more appropriate than no protest. Most notably, these positive feelings for a protestor
and her response were moderated by perceptions of pervasive gender discrimination. As beliefs about pervasive gender
discrimination increased, so did liking for the protestor and appropriateness ratings of the protest response, whereas anger
towards the target decreased. The opposite occurred for a non-protestor: as pervasiveness beliefs increased, a nonprotestor
elicited greater anger, less liking and lower appropriateness ratings. Together these ﬁndings support our
predictions that beliefs about gender inequality (Tajfel & Turner, 1986) shape how women perceive ingroup responses to
experiences with discrimination. Our ﬁndings suggest that women who perceive gender discrimination as pervasive will be
positive towards a protest response and protestor, and negative when protest is absent.
Our prediction concerning the mediational role of response appropriateness on liking and anger was also supported.
Perceptions of gender discrimination as pervasive were associated with increased beliefs about the appropriateness of a
protest response and decreased beliefs about the appropriateness of a non-protest. These beliefs about response
appropriateness generated greater positive affective and evaluative responses towards a protestor, but greater negative
reactions towards a non-protestor. These ﬁndings suggest that women who regarded their group as pervasively
disadvantaged preferred a protestor because they regarded her response as being more appropriate than no protest.
Paralleling past research showing that people are supportive of complaints (Kowalski, 1996) and collection actions
(Hornsey et al., 2006; Wright et al., 1990) that they perceive as warranted, the present research indicates that women’s
responses to ingroup discrimination claimers may depend upon the perceived appropriateness of the complaint.
The type of protest chosen by the target did not produce any effects on anger or liking for the target. Women’s proclivity
to prefer a target who protested the hiring decision did not vary as a function of whether she employed an individual or
collective protest strategy. Moreover, although beliefs about the pervasiveness of discrimination moderated evaluations of
a protestor versus non-protestor, these beliefs did not produce a similar moderation effect for the type of protest. Thus,
target evaluations were unaffected by protest type, regardless of women’s beliefs about their collective disadvantage.
Appropriateness ratings did differ; however, across the two types of protest. There was a marginal tendency for women to
regard individual protest as more appropriate than collective protest. The results for response appropriateness; however,
were limited to this main effect.
The ﬁnding that participants favoured a target who protested the hiring decision, regardless of whether she employed an
individual or collective protest strategy appears counterintuitive, especially when applied to women who strongly believe
that their group experiences pervasive discrimination. These women could be expected to prefer strategies that stress
women’s collective assets rather than an individual woman’s ability. So, why did we not ﬁnd that evaluations of the female
protestor varied as a function of the form of protest she took in conjunction with beliefs about pervasive discrimination?
To address the above question, we draw on both the concern about women measure and the appropriateness ratings of
the different forms of protest. Recall that women in our study regarded (1) both the individual and collective protestors to
be concerned about women as a group relative to a non-protestor, and (2) the individual protest as more appropriate than
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
742 Donna M. Garcia et al.
the collective one. These ﬁndings are consistent with research showing that in Western societies, people generally prefer
individual rather than collective protest (Ellemers, Wilke, & van Knippenberg, 1993), even when it is clear that token
mobility is possible for only a few individuals (see Wright, 2001). Because individual approaches to employment mobility
are congruent with workplaces that emphasize individual responsibility for personal outcomes, they may be especially
favoured in these settings. The tendency for our participants to be more supportive of an individual strategy could reﬂect
both beliefs about the strategy’s efﬁcacy in the given context and the inference that the female lawyer was using similar
knowledge to be strategic. Women may fear that in an individualistic ‘boys’ club’ (such as a law ﬁrm) publicly claiming
discrimination will be socially undesirable and result in further harm (see Kobrynowicz & Branscombe, 1997). A protest
that insinuates that men are sexist and disregard women’s abilities would be more threatening to decision makers than one
that stresses an individual woman’s abilities and does not directly challenge the men’s views about women. If an individual
strategy was regarded by participants as minimizing potential defensiveness in the men, it could then be considered more
effective than arguing ‘women’s cause’ at enhancing the woman’s future success in the ﬁrm. In addition, regardless of the
strategy used, individual women who achieve advancement might be seen as opening doors for other women, and
improving women’s status overall. Indeed, women in our study did regard an individualistic protestor as being concerned
about women as a group.
The preference for the individual strategy could also be attributed to the inexperience of our participants, who may be
unaware that individual mobility approaches are successful for only a few women and not foresee how such approaches
could undermine women’s interests (Wright, 2001). Indeed, our ﬁnding that women regarded an individual strategy as
showing concern for women as a group (relative to a non-protestor) suggests they did not realize the harmfulness of that
approach for their collective interests. Unfortunately, the preference to adopt individual strategies to improve personal
status might be further encouraged if such behaviour elicits appreciation from ingroup members. Yet, the use of individual
strategies tends to be ineffective at fostering positive change and undermines women’s support for collective action (see
Wright, 2001). Thus, our participants’ support for a woman using individual mobility strategies can ironically be
detrimental to women, individually and collectively.
Implications, Limitations and Future Directions
The mediational role of perceived appropriateness raises some caution regarding the empirical investigation of the social
consequences faced by discrimination claimers. The type of complaint made by discrimination claimers often differs
across research studies. In some research, participants learn that after failing a test graded by a potentially biased evaluator,
the target attributed the failure to discrimination in a survey about the test-taking experience (e.g. Garcia et al., 2005;
Kaiser & Miller, 2001). In other studies, the participant learn that the target confronted the discriminator (e.g. Dodd et al.,
2001). As argued by Garcia et al. (2005), circling a number on an anonymous survey is not equivalent to confronting the
perpetrator or protesting the outcome. The latter two are likely to be perceived by participants as a more effective means
than the former and research shows that people will be supportive of complaints they deem as having instrumental value
(Kowalski, 1996). Despite these important differences, the various responses to discrimination are often used
interchangeably within the literature, and the import of the particular type of discrimination claim used in a speciﬁc study
is seldom discussed. One fruitful future research direction would be to examine directly the consequences associated with
various forms of discrimination claims.
There are other methodological features worth considering that provide insight into discrepant ﬁndings across past
research studies. For example, we found that women expressed less anger and greater liking toward a woman who
protested rather than did not protest the promotion decision. These ﬁndings are in contrast to the ingroup social costs
observed by Garcia et al. (2005) but are consistent with the ingroup support observed by Dodd et al. (2001) in the blatant
sexism condition. Given that participants strongly agreed that the female lawyer experienced discrimination, the event in
the current study more closely resembled the blatant condition in the Dodd et al. study. With this similarity, our ﬁndings
expand upon those by Dodd et al. by showing that even when discrimination is blatant, subjective beliefs about gender
discrimination in society as a whole (not just the objective likelihood of discrimination in a speciﬁc situation) can inﬂuence
support for ingroup discrimination claimers.
The current study is similar to Dodd et al. in two other key ways. First, like Dodd et al., the current study used a
discriminatory outcome that was more consequential than the one used by Garcia et al. (2005). Being refused a deserved
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
Women’s reactions to ingroup protestors 743
promotion in a prestigious law ﬁrm because of one’s gender is more consequential than failing a test in a research study.
People tend to be more supportive of complaints against consequential rather than trivial outcomes (Kowalski, 1996).
Furthermore, as noted earlier, protesting the outcome (current study) or confronting a man for his sexist comment (Dodd
et al., 2001) is a more effective means of addressing the issue than is privately circling a survey response (Garcia et al.,
2005).
The possibility that different research approaches contribute to different ﬁndings in the social costs literature provides
an opportunity to expand our understanding of the factors that inﬂuence reactions to discrimination claimers. In real world
settings, discrimination claims can take on several different forms and address a variety of discriminatory outcomes, which
vary in ambiguity and consequentiality. Based on the discrepant ﬁndings across the research discussed above, we propose
that support for ingroup discrimination claimers will be greatest when the discrimination is consequential and
unambiguous, and the discrimination claim could foster positive change (see Garcia et al., 2005). Although our research
indicates that appropriateness does matter, it would be informative to isolate what factors lead to perceptions that protest is
appropriate. We suggest that manipulating the effectiveness of the response (the discrimination claim results in positive,
negative, or no change) would be a useful direction for further research.
In summary, the current research extends past ﬁndings that publicly claiming discrimination can have social
consequences by examining two factors that we hypothesized would inﬂuence women’s evaluations of a discrimination
claimer: beliefs about the pervasiveness of gender discrimination and perceptions about the appropriateness of the protest.
We argued that women’s reactions to a woman who protested an experience with gender discrimination would be
determined by the perceived implications that the protestor has for the ingroup. If the woman’s response was perceived as
harming the ingroup, she would face ingroup rejection. If her protest was regarded as beneﬁting the ingroup, she would
experience support. Consistent with this argument we found that women who believed their group to be targeted by gender
discrimination were more supportive of a woman who protested rather than accepted a discriminatory outcome. Further,
the joint effects of response type (protest versus no protest) and perceptions of pervasive sexism were mediated by
response appropriateness. Perceiving pervasive sexism led women to regard protest as more appropriate than no protest,
which in turn generated positive evaluations of a protestor and negative reactions to a non-protestor.
CONCLUSION
Past research indicates that people will be especially inclined to tailor their public responses to discrimination in order to
maximize ingroup approval (e.g. Ellemers et al., 2004; Postmes & Branscombe, 2002). Given the many advances that
groups have made as a result of individual and collective protest, it is essential for a group’s well being that people feel
encouraged by their own ingroup to confront discrimination and use collective mobility strategies. Although silence and
individual mobility approaches may provide temporary beneﬁts for the individual and group’s social identity—
particularly if nothing effective can be done in the particular circumstances, it can also prevent long-term progress.
Consequently, further investigation is warranted to understand the factors that inﬂuence the social forces that shape how
people respond to personal experiences with discrimination.
REFERENCES
Adams, G., Garcia, D. M., Purdie-Vaughns, V., & Steele, C. M. (2006). The detrimental effects of a suggestion of sexism in an
instruction situation. Journal of Experimental Social Psychology, 42, 602–615.
Branscombe, N. R., & Ellemers, N. (1998). Coping with group-based discrimination: Individualistic versus group-level strategies. In
J. K. Swim & C. Stangor (Eds.), Prejudice: The target’s perspective (pp. 243–266). San Diego, CA: Academic Press.
Branscombe, N. R., Wann, D. L., Noel, J. G., & Coleman, J. (1993). In-group or out-group extremity: Importance of the threatened social
identity. Personality and Social Psychology Bulletin, 19, 381–388.
Charles, M., & Grusky, D. B. (2004). Occupational ghettos: The worldwide segregation of women and men. Palo Alto, CA: Stanford
University Press.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analyses for the behavior sciences (3rd
ed.). Mahwah, NJ: Lawrence Erlbaum Associates.
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
744 Donna M. Garcia et al.
Dixon, K. A., Storen, D., & Horn, C. E. V. (2002). A workplace divided: How Americans view discrimination and race on the job,
Rutgers, The State University of New Jersey: John J. Heldrich Center for Workforce Development.
Dodd, E. H., Giuliano, T. A., Boutell, J. M., & Moran, B. E. (2001). Respected or rejected: Perceptions of women who confront sexist
remarks. Sex Roles, 45, 567–577.
Ellemers, N. (1993). The inﬂuence of socio-structural variables on identity management strategies. In W. Stroebe & M. Hewstone (Eds.),
European Review of Social Psychology (Vol. 4, pp. 27–58). New York: Wiley.
Ellemers, N., Doosje, B., & Spears, R. (2004). Sources of respect: The effects of being liked by ingroups and outgroups. European
Journal of Social Psychology, 34, 155–172.
Ellemers, N., Wilke, H., & Van Knippenberg, A. (1993). Effects of the legitimacy of low group or individual status on individual and
collective status-enhancement strategies. Journal of Personality and Social Psychology, 64, 766–778.
Feagin, J. R., & Sikes, M. P. (1994). Living with racism: The Black middle-class experience. Boston, MA: Beacon.
Garcia, D. M., Horstman Reser, A., Amo, R., Redersdorff, S., & Branscombe, N. R. (2005). Perceivers’ responses to in-group and outgroup
members who blame a negative outcome on discrimination. Personality and Social Psychology Bulletin, 31, 769–780.
Gupta, V. K., & Bhawe, N. M. (2007). The inﬂuence of proactive personality and stereotype threat on women’s entrepreneurial
intentions. Journal of Leadership and Organizational Studies, 13, 73–85.
Hornsey, M. J., Blackwood, L., Louis, W., Fielding, K., Mavor, K., Morton, T., et al. (2006). Why do people engage in collective action?
Revisiting the role of effectiveness. Journal of Applied Social Psychology, 36, 1701–1722.
Kaiser, C. R., Dyrenforth, P. S., & Hagiwara, N. (2006). Why are attributions to discrimination interpersonally costly? A test of system
and group justifying motivations. Personality and Social Psychology Bulletin, 32, 1523–1536.
Kaiser, C. R., & Miller, C. T. (2001). Stop Complaining! The social costs of making attributions to discrimination. Personality and
Social Psychology Bulletin, 27, 254–263.
Kobrynowicz, D., & Branscombe, N. R. (1997). Who considers themselves victims of discrimination? Individual difference predictors
of perceived gender discrimination in women and men. Psychology of Women Quarterly, 21, 347–363.
Kowalski, R. M. (1996). Complaints and complaining: Functions, antecedents, and consequences. Psychological Bulletin, 119, 179–
196.
Major, B., & O’Brien, L. T. (2005). The social psychology of stigma. Annual Review of Psychology, 56, 393–421.
Marques, J. M., Yzerbyt, V. Y., & Leyens, J.-P. (1988). The ‘‘black sheep effect’’: Extremity of judgments towards in-group members as
a function of group identiﬁcation. European Journal of Social Psychology, 18, 1–16.
Morgan-Lopez, A. A., & MacKinnon, D. P. (2006). Demonstration and evaluation of a method for assessing mediated moderation.
Behavior Research Methods, 38, 77–87.
Muller, D., Judd, C. M., & Yzerbyt, V. Y. (2005). When moderation is meditated and when mediation is moderated. Journal of
Personality and Social Psychology, 89, 852–863.
Postmes, T., & Branscombe, N. R. (2002). Inﬂuence of long term racial environmental composition on subjective well being in African
Americans. Journal of Personality and Social Psychology, 83, 735–751.
Postmes, T., Branscombe, N. R., Spears, R., & Young, H. (1999). Comparative processes in personal and group judgments: Resolving
the discrepancy. Journal of Personality and Social Psychology, 76, 320–338.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple
mediator models. Behavior Research Methods, 40, 879–891.
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation hypotheses: Theory, methods, and prescriptions.
Multivariate Behavioral Research, 42, 185–227.
Schmitt, M. T., & Branscombe, N. R. (2002). The meaning and consequences of perceived discrimination in disadvantaged and
privileged social groups. European Review of Social Psychology, 12, 167–199.
Simon, B., Hastedt, C., & Aufderheide, B. (1997). When self-categorization makes sense: The role of meaningful social categorization
in minority and majority members’ self-perception. Journal of Personality and Social Psychology, 73, 310–320.
Steele, C. M., Spencer, S. J., & Aronson, J. (2002). Contending with group image: The psychology of stereotype and social identity
threat. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 379–440). San Diego, CA: Academic Press.
Stangor, C., Swim, J. K., Van Allen, K. L., & Sechrist, G. B. (2002). Reporting discrimination in public and private contexts. Journal of
Personality and Social Psychology, 82, 69–74.
Swim, J. K., & Hyers, L. L. (1999). Excuse me-what did you say? Women’s public and private responses to sexist remarks. Journal of
Experimental Social Psychology, 35, 68–88.
Tajfel, H. (1978). Differentiation between social groups: Studies in the social psychology of intergroup relations. London: Academic
Press.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel & W. G. Austin (Eds.), The social
psychology of intergroup relations (2nd ed. pp. 7–24). Monterey, CA: Brooks-Cole.
Wright, S. C. (2001). Restricting intergroup boundaries: Tokenism, ambiguity, and the tolerance of injustice. In J. T. Jost & B. Major
(Eds.), The psychology of legitimacy (pp. 223–256). Cambridge, UK: Cambridge University Press.
Wright, S. C., Taylor, D. M., & Moghaddam, F. M. (1990). Responding to membership in a disadvantaged group: From acceptance to
collective protest. Journal of Personality and Social Psychology, 58, 994–1003.
Copyright # 2009 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 40, 733–745 (2010)
DOI: 10.1002/ejsp
Women’s reactions to ingroup protestors 745