Content uploaded by Sapna Cheryan
Author content
All content in this area was uploaded by Sapna Cheryan on Apr 06, 2017
Content may be subject to copyright.
Do Female and Male Role Models Who
Embody STEM Stereotypes Hinder
Women’s Anticipated Success in STEM?
Sapna Cheryan
1
, John Oliver Siy
1
, Marissa Vichayapai
1
,
Benjamin J. Drury
1
, and Saenam Kim
1
Abstract
Women who have not yet entered science, technology, engineering, and mathematics (STEM) fields underestimate how well they
will perform in those fields (e.g., Correll, 2001; Meece, Parsons, Kaczala, & Goff, 1982). It is commonly assumed that female role
models improve women’s beliefs that they can be successful in STEM. The current work tests this assumption. Two experiments
varied role model gender and whether role models embody computer science stereotypes. Role model gender had no effect on
success beliefs. However, women who interacted with nonstereotypical role models believed they would be more successful in
computer science than those who interacted with stereotypical role models. Differences in women’s success beliefs were
mediated by their perceived dissimilarity from stereotypical role models. When attempting to convey to women that they can
be successful in STEM fields, role model gender may be less important than the extent to which role models embody current
STEM stereotypes.
Keywords
stereotypes, role models, gender, STEM, anticipated success
Women are less likely than men to enter science, technology,
engineering, and mathematics (STEM) fields (Hill, Corbett,
& St. Rose, 2010)—a disparity that exists even when men and
women are matched for quantitative ability and experience
(Strenta, Elliott, Adair, Matier, & Scott, 1994). An important
precursor to entering a field is anticipating success in it
(Bandura, 1997; Wigfield & Eccles, 2000). An explanation put
forth for the gender disparity in STEM participation is that
women tend to underestimate their abilities to be successful
in these fields (Correll, 2001; Ehrlinger & Dunning, 2003;
Meece, Parsons, Kaczala, & Goff, 1982; Miura, 1987; Sax,
1994). One common way to convey to women that they can
be successful in STEM is to expose them to a STEM role
model, or someone who is successful in these fields and can
be emulated (Lockwood & Kunda, 1997; Marx, Stapel, &
Muller, 2005). It is widely accepted that female role models are
more effective than male role models in inspiring girls and
women to enter STEM fields (see Mattel’s Computer Engineer
Barbie and the National Academy of Sciences’ book series
Women’s Adventures in Science). Yet, even as efforts to intro-
duce women to female role models have become increasingly
widespread, the proportion of women who pursue computer
science and math at the college level has remained stagnant
or even decreased over the past few decades (National Science
Foundation, 2009). Are female role models more effective than
male role models in encouraging women who are not already in
STEM to believe they can succeed in these fields? We propose
that when conveying to women their potential for future suc-
cess, role model gender may be less important than the extent
to which role models embody current STEM stereotypes.
Male-dominated fields can be unwelcoming to women on
two dimensions. The first is gender ratio, or the extent to which
there are more men than women in a field. A skewed gender
ratio can activate negative stereotypes about women’s abilities
and bring about underperformance among women who have
identified those domains as important to them (e.g., STEM
majors; Inzlicht & Ben-Zeev, 2000; Murphy, Steele, & Gross,
2007; Sekaquaptewa & Thompson, 2003). Female role models
protect women who are personally invested in STEM against
the harmful effects of gender stereotypes by preventing
underperformance (Marx & Roman, 2002), improving their
self-views (Lockwood, 2006; Marx & Roman, 2002), and
improving their implicit attitudes toward the domain (Stout,
1
Department of Psychology, University of Washington, Seattle, WA, USA
Corresponding Author:
Sapna Cheryan, Department of Psychology, University of Washington, Box
351525, Seattle, WA 98195, USA
Email: scheryan@uw.edu
Social Psychological and
Personality Science
000(00) 1-9
ªThe Author(s) 2011
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1948550611405218
http://spps.sagepub.com
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
Dasgupta, Hunsinger, & McManus, 2011). However, concerns
about negative gender stereotypes are less threatening to
women who are not personally invested in the domain (Schma-
der, Johns, & Forbes, 2008; Steele, 1997). Indeed, among a
sample of noncomputer science majors, perceived dissimilarity
from computer science majors better predicted women’s lack
of interest in computer science than their concerns about neg-
ative gender stereotypes and their estimated percentage of men
in computer science (Cheryan & Plaut, 2010; see also Cher-
yan, Plaut, Davies, & Steele, 2009; Walton & Cohen, 2007 for
belonging threats in the absence of stereotype threat effects).
Thus, deploying female instead of male role models in com-
puter science, an intervention to counteract effects of negative
gender stereotypes, may be less effective among our popula-
tion of interest, namely women who are not already personally
invested in these fields.
The second dimension of male-dominated fields that is
unwelcoming to women is the extent to which the field is
assumed to embody stereotypes that are incongruous with the
female gender role (Cheryan et al., 2009; Diekman, Brown,
Johnston, & Clark, 2010). In STEM, these stereotypes include
a tendency toward social isolation and a singular focus on tech-
nology (Barbercheck, 2001). Computer scientists in particular
are stereotyped as ‘‘computer nerds’’ who are socially awkward
and obsessed with computers (Margolis & Fisher, 2002; Schott
& Selwyn, 2000). In contrast, the female gender role prescribes
many opposing characteristics—helping and working with oth-
ers, being socially skilled, and attending to physical appearance
(Cejka & Eagly, 1999; Diekman et al., 2010; Eagly & Steffen,
1984). Gender roles shape the way people see themselves
(Eagly, 1987), and women report feeling dissimilar from people
who fit STEM stereotypes (Cheryan et al., 2009). Feeling dis-
similar from others causes people to contrast their self-views
away from them, or believe themselves less likely to possess
traits that others have (Brown, Novick, Lord, & Richards,
1992; Mussweiler, 2001, 2003). Women who encounter stereo-
typical STEM students may feel dissimilar from them and, as a
result, underestimate their likelihood of succeeding in STEM. As
a result, STEM-stereotypical role models who are supposed to
inspire emulation may backfire and discourage those they were
meant to benefit.
Though STEM stereotypes are incongruent with the female
gender role, we propose that they can be conveyed by women
as well. Examples abound in our society of women embodying
characteristics that are incongruent with the female gender role
(e.g., tomboys; Deaux & Lewis, 1984; Fagot, 1977; Rudman &
Fairchild, 2004). When women embody STEM stereotypes,
they may evoke in other women feelings of dissimilarity and
cause contrasted self-views, despite their shared gender.
Indeed, when presented with information about another per-
son’s gender and his or her gendered characteristics (e.g.,
sturdy, broad-shoulders), inferences about that person are
based more on his or her gendered characteristics than on gen-
der (Deaux & Lewis, 1984). When female role models embody
these stereotypical traits, they may be just as powerful of a
deterrent as when males embody them.
Overview and Hypotheses
In two studies, we investigate how gender and stereotypicality
of role models influence success beliefs in computer science
among a population of women who are not already in the field.
In contrast to previous work that manipulated whether or not
the role model was in the domain (Lockwood & Kunda,
1997; Marx & Roman, 2002; Marx et al., 2005) or perceived
as competent (Buunk, Peiro´, & Griffioen, 2007; Lockwood,
Marshall, & Sadler, 2005; Marx & Roman, 2002; Marx et al.,
2005), we vary stereotypes associated with the people in the
domain (e.g., liking science fiction) while keeping STEM
membership and perceived competence constant. This is
important because it would demonstrate that even role models
who are competent in STEM can hinder women’s anticipated
success in STEM to the extent that they embody STEM
stereotypes.
We hypothesize that encountering a stereotypical computer
science role model, irrespective of the role model’s gender, will
decrease women’s—but not men’s—success beliefs in com-
puter science compared to encountering a nonstereotypical
computer science role model or no role model. Moreover, we
predict that feelings of dissimilarity to stereotypical role mod-
els will mediate women’s decreased success beliefs. Finally,
we predict that role model gender will have less of an influence
on women’s beliefs that they can be successful in STEM than
role model stereotypicality. Given that women’s underrepre-
sentation in STEM is more attributable to inadequate recruit-
ment than inadequate retention (Ceci, Williams, & Barnett,
2009; de Cohen & Deterding, 2009), understanding the factors
that prevent women who are not already personally invested in
STEM from feeling like they can succeed in STEM will be cru-
cial to remedying gender disparities (Bandura, 1997; Betz &
Hackett, 1981; Ehrlinger & Dunning, 2003; Miura, 1987;
Wigfield & Eccles, 2000).
Study 1
Study 1 investigated whether interacting with a computer
science role model influences women’s success beliefs in com-
puter science. To ensure role models would be relatable to our
participants and their success in STEM attainable (Lockwood
& Kunda, 1997), we used upper-level undergraduates as role
models. This is similar to previous research (Lockwood
et al., 2005; Marx & Roman, 2002; Stout et al., 2011) and to
situations in which upper-level undergraduates are chosen to
be role models for other undergraduates, for example, as teach-
ing assistants and resident advisors. Stereotypicality of role
models was manipulated using pretesting clothing, hobbies,
and preferences that are associated with computer science
majors (stereotypical) or average college students (nonstereo-
typical). Nonstereotypical role models were modeled on aver-
age college students, rather than on a more extreme stereotype
violator, to reduce the likelihood that they would be subtyped
as unrepresentative of the field (see Kunda & Oleson, 1995).
This study also included a baseline condition (no role model)
2Social Psychological and Personality Science 000(00)
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
to examine which condition drove effects. Finally, we tested
for a potential alternative explanation: that stereotypical role
models would arouse feelings of gender-based threat and there-
fore deter women from computer science (see Davies, Spencer,
Quinn, & Gerhardstein, 2002).
Method
Participants were 85 female noncomputer science majors from
the psychology participant pool. Noncomputer science majors
were used to focus on recruitment. Two participants were elim-
inated because they did not remember the confederate’s major.
Pretest
Students (N¼31; 22 women) were asked to list clothing and
hobbies that they associate with computer science majors
(stereotypical) and college students (nonstereotypical). Fre-
quently listed clothing was consolidated to make the four out-
fits (i.e., male stereotypical, male nonstereotypical, female
stereotypical, and female nonstereotypical), and the three most
frequently listed hobbies in each category were selected
1
(see
Table 1). To pretest the selected outfits, full-length photos of
the outfits were superimposed onto male or female stick
figures and rated by a separate sample (N¼22; 12 women) for
how much each outfit fits the stereotype of a computer science
major, on a scale from 1 (not at all)to7(extremely). Partici-
pants saw all four outfits, and order of presentation of outfits
was counterbalanced. A 2 (Stereotypicality) 2 (Outfit Gen-
der) 2 (Participant Gender) mixed-model analysis of var-
iance (ANOVA) revealed the predicted main effect of outfit,
F(1, 20) ¼67.82, p< .001, Z2
p¼.77. Stereotypical outfits were
rated as significantly more stereotypical (M¼5.61, SD ¼.91)
than nonstereotypical outfits (M¼3.34, SD ¼1.13). There
were no other main effects or interactions. Two separate
samples of students also rated how much they associated the
selected preferences (i.e., favorite movie, magazine, and
television show; see Table 1; N¼27; 13 women) and hobbies
(N¼33; 19 women) with computer science majors. A 2
(Stereotypicality) 2 (Participant Gender) ANOVA on prefer-
ences (averaged together) and another identical ANOVA on
hobbies revealed that the stereotypical items were significantly
more associated with computer science majors (preferences:
M¼4.40, SD ¼1.45; hobbies: M¼5.39, SD ¼1.22) than the
nonstereotypical items (preferences: M¼2.43, SD ¼1.02;
hobbies: M¼3.24, SD ¼.96), both Fs > 62, ps < .001,
Z2
ps > .69. There were no other main effects or interactions.
Procedure
Participants interacted with one of four confederates (two White
females, two White males) who posed as a fellow participant.
Participants and confederates engaged in a ‘‘getting to know
each other’’ task, which comprised of a list of printed questions
to ask each other. The first four questions were answered by the
confederates the same way regardless of stereotypicality: what is
your first name (‘‘Jennifer’’ or ‘‘David’’), what year are you in
school (‘‘junior’’), what are you majoring in (‘‘computer sci-
ence’’), and where are you from (‘‘Seattle’’). Stereotypicality
was manipulated via confederate clothing (see Table 1; photos
available upon request) and answers to the last four questions:
what are your hobbies, what is your favorite movie, what is your
favorite television show, and what is your favorite magazine (see
Table 1). Participants were randomly assigned to stereotypicality
and gender conditions. All four confederates performed both
stereotypical and the nonstereotypical conditions, and confeder-
ates were trained to have identical nonverbal behaviors and ver-
bal fillers (e.g., ‘‘I like ...’’) across conditions. Order of who
asked questions first was counterbalanced. Interactions lasted
on average 1 min 41 s.
After the interaction, participants were separated from
confederates and filled out a questionnaire in which they
recalled their partner’s responses. Success beliefs were
assessed using two questions after they recalled their partner’s
major:‘‘Howwelldoyouthinkyouwoulddomajoringinthat
field?’’ and ‘‘How well would you perform as someone who is
a major in that field?’’ on scales from 1 (notwellatall)to7
(very well); adapted from Meece, Wigfield, and Eccles
(1990), r¼.82, p< .001.
To investigate dissimilarity as a potential mediator of the
effect, participants were asked how similar they were to their
partner on a scale from 1 (not at all)to7(very much). We also
included questions on other potential mediators, including how
well they got along with their partner, on a scale from 1 (not at
all)to7(very well), and five questions assessing gender-based
threat (e.g., ‘‘How anxious would you be about confirming a
negative stereotype about your gender if you majored in that
field?’’ a¼.92; adapted from Cohen & Garcia, 2005; Marx
et al., 2005), on a scale from 1 (not at all)to7(extremely).
Table 1. Stereotypical and Nonstereotypical Items in Both Studies
Stereotypical Nonstereotypical
Clothing Glasses, a t-shirt that read ‘‘I code therefore
I am,’’ unfashionable pants, socks and sandals
Solid-colored shirt (v-neck t-shirt for
women, polo for men), jeans, flip-flops
Hobbies Playing video games, watching anime, and programming Playing sports, hanging out with
friends, and listening to music
Favorite movie Star Wars American Beauty
Favorite television show Mystery Science Theater 3000 The Office
Favorite magazine Electronic Gaming Monthly Rolling Stone
Cheryan et al. 3
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
Participants also indicated how intelligent they found their
partner, on a scale from 1 (not at all)to7(very much).
2
Some participants were randomly assigned to a baseline
condition where they found out that their partner (gender
unspecified) did not show up. They then drew slips to find out
what major to rate (all slips said ‘‘computer science’’). This
study thus employed a 2 (Stereotypicality) 2 (Role Model
Gender) between-subjects design, with an additional baseline
condition. The study concluded with demographic questions.
Results and Discussion
Success Beliefs
A 2 (Stereotypicality) 2 (Role Model Gender) ANOVA
revealed a main effect of stereotypicality, F(1, 55) ¼8.31,
p< .01, Z2
p¼.13. Women anticipated lower success in com-
puter science after briefly interacting with a stereotypical
(M¼2.18, SD ¼1.08) compared to a nonstereotypical
(M¼3.09, SD ¼1.32) role model. There was no main effect
of role model gender and no interaction, both Fs<1,ns.
Comparison to Baseline
To examine baseline participants, we conducted a one-way
ANOVA on success beliefs, collapsed across role model gen-
der (because baseline participants did not encounter a role
model), which revealed an effect of condition, F(2, 80) ¼4.61,
p< .05 (see Figure 1). Tukey’s post hoc comparisons revealed
that women in the baseline condition anticipated greater suc-
cess (M¼3.10, SD ¼1.62) than those in the stereotypical
condition (M¼2.18, SD ¼1.08), p<.05,d¼.67, and a
similar level of success to those in the nonstereotypical condi-
tion (M¼3.09, SD ¼1.32), ns. Computer science role models
who embody stereotypes of computer science are thus detrimental
to women’s anticipated success in computer science.
Perceived Dissimilarity
A 2 (Stereotypicality) 2 (Role Model Gender) ANOVA on
similarity revealed a main effect of stereotypicality, F(1, 54) ¼
14.27, p<.001,Z2
p¼.21. Women rated themselves as signifi-
cantly less similar to stereotypical (M¼2.67, SD ¼1.52) than
nonstereotypical (M¼4.00, SD ¼1.12) role models. There
was also a weak marginal main effect of role model gender
such that women rated themselves as marginally more similar
to female (M¼3.63, SD ¼1.31) thanmale (M¼3.03, SD ¼1.60)
role models, F(1, 54) ¼2.88, p¼.10, Z2
p¼.05. Stereotypicality
and role model gender did not interact, F<1,ns.
Success Beliefs Mediated by Perceived Dissimilarity
We examined whether the relationship between stereotypical-
ity and success beliefs in computer science was mediated by
perceived dissimilarity to stereotypical role models using the
steps outlined by Baron and Kenny (1986) and the SPSS macro
developed by Preacher and Hayes (2004), with 5000 bootstrap-
ping resamples. In Steps 1 and 2, as seen above, compared to
nonstereotypical role models, interacting with a stereotypical
role model decreased success beliefs, b¼–.87, SE ¼.31,
p< .01, and decreased feelings of similarity, b¼–1.33,
SE ¼.35, p< .001. In Steps 3 and 4, similarity predicted
success beliefs upon controlling for stereotypicality, b¼.34,
SE ¼.11, p< .01, and stereotypicality was no longer related
to success beliefs, b¼–.42, SE ¼.33, ns; Sobel Z¼2.32,
p< .05 (95%CI: [–.82, –.14]). Thus, perceptions of dissimilar-
ity to stereotypical role models compared to nonstereotypical
role models accounted for women’s lower success beliefs in
computer science.
Alternative Explanations
A 2 (Stereotypicality) 2 (Role Model Gender) ANOVA on
gender-based threat revealed no main effects or interactions,
F(1, 55)s < 1.28, ps > .26, suggesting that interacting with
stereotypical role models did not arouse such concerns com-
pared to interacting with nonstereotypical role models. A 2
(Stereotypicality) 2 (Role Model Gender) ANOVA on how
well participants got along with their partner also revealed no
main effects or interactions, F(1, 54)s < 2.23, ps > .14, suggest-
ing that confederates did not inadvertently change their interac-
tion styles across conditions. To make certain that
confederates’ behaviors were constant across conditions, our
next study used a different method that afforded complete con-
trol over confederates’ behaviors.
Study 2
Study 2 assessed how computer science role models influence
both women’s and men’s success beliefs. We hypothesized that
1
2
3
4
baseline stereotypical nonstereotypical
Success beliefs in CS
(1 to 7 scale)
Figure 1. Women’s beliefs about whether they will be successful in
computer science (CS) after interacting with no role model (baseline),
a stereotypical computer science role model, or a nonstereotypical
computer science role model in Study 1.
4Social Psychological and Personality Science 000(00)
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
because the incompatibility between computer science stereo-
types and gender role is greater for females than males
(Cheryan et al., 2009; Diekman et al., 2010), women’s suc-
cess beliefs would be affected by the stereotypical role model
but men’s success beliefs would be unaffected. Assessing
effects on men is theoretically important because it reveals
whether the process by which STEM stereotypes deter
women is, as we predict, a gendered one, or whether there
is something threatening or off-putting about the stereotypi-
cal role model to both women and men. We also moved
Study 2 to a virtual environment, which enabled complete
standardization of role models’ behaviors across condition.
Previous work has found that people act in virtual environ-
ments in accordance with their real-life social identities
(Dotsch & Wigboldus, 2008; Eastwick & Gardner, 2009;
McKenna & Bargh, 1998), and both men and women are
influenced by interactions in virtual environments (e.g.,
Okita, Bailenson, & Schwartz, 2007).
Method
Participants included noncomputer science majors (N¼88) in
the participant pool. Eight participants were eliminated for
suspicion, six were eliminated for technical difficulties
(e.g., avatar did not load properly), five were eliminated for
living in the United States for 1 year or less because computer
science stereotypes are different abroad (Othman & Latih,
2006; Varma, 2009), and one was eliminated for not remem-
bering the role model’s major.
3
A total of 68 participants
(40 women) remained.
Participants were told that the study investigated ‘‘how a
virtual world can be used as a tool for getting to know some-
one else.’’ Participants learned that they would have a virtual
interaction with another student (actually a confederate)
located in another part of the building. Participants’ photos
were taken and ostensibly superimposed onto an avatar to
‘‘personalize their virtual self.’’ This was to ensure that par-
ticipants saw their partner’s avatar as a representation of their
partner’s appearance. Participants were logged into Second
Life, an online 3D virtual environment (http://secondlife.
com). First-person view was utilized so that participants’ own
avatars were not visible. After completing a short tutorial on
Second Life, participants navigated their avatars into a small
room and sat down across from their partner’s avatar. Stereo-
typicality of partner’s avatar was manipulated using the same
clothing (digitally represented; photos available upon
request), hobbies, and stated preferences as Study 1. Gender
of avatar was also manipulated.
Participants and confederates engaged in the same ‘‘getting to
know each other’’ task used in Study 1, with identical confeder-
ate answers adapted for online chatting (e.g., ‘‘Oh ... Ialso
really like listening to music’’). Typed questions and responses
appeared as text on screen. After interactions were complete,
participants answered the same questions from Study 1 on suc-
cess beliefs, r¼.70, p< .001, and similarity to their partner.
4
Results and Discussion
Success Beliefs
A 2 (Stereotypicality) 2 (Participant Gender) 2 (Role
Model Gender) ANOVA on success beliefs revealed a main
effect of participant gender, F(1, 60) ¼13.69, p< .001,
Z2
p¼.19. Women reported lower success beliefs in computer
science (M¼2.45, SD ¼1.17) than did men (M¼3.59,
SD ¼1.38). This main effect was qualified by the predicted
Stereotypicality Participant Gender interaction, F(1, 60) ¼
4.65, p< .05, Z2
p¼.07 (see Figure 2). For women, a brief, vir-
tual interaction with a stereotypical role model lowered success
beliefs in computer science (M¼2.00, SD ¼.96) compared to
an interaction with a nonstereotypical role model (M¼2.95,
SD ¼1.19), p< .05, F(1, 60) ¼5.57, p< .05, Z2
p¼.09. Men’s
success beliefs were unaffected by interacting with either
the stereotypical (M¼3.82, SD ¼1.07) or nonstereotypical
(M¼3.36, SD ¼1.65) role models, F(1, 60) < 1, ns. There
were no other main effects or interactions. That men’s success
beliefs were not affected by exposure to stereotypical role mod-
els reveals that these stereotypes communicate to women, but
not to men, a lower potential for success in the field.
Perceived Dissimilarity
A 2 (Stereotypicality) 2 (Participant Gender) 2 (Role
Model Gender) ANOVA on similarity to role models revealed
a main effect of stereotypicality, F(1, 60) ¼8.55, p<.01,Z2
p¼.13.
Participants felt less similar to stereotypical (M¼2.89,
SD ¼1.39) than nonstereotypical role models (M¼3.91,
SD ¼1.04). There was also a marginal main effect of
Participant Gender, F(1, 60) ¼3.58, p¼.06, Z2
p¼.06, such
that men reported more similarity (M¼3.68, SD ¼1.09) than
women (M¼3.18, SD ¼1.45). These main effects were qual-
ified by the predicted Stereotypicality Participant Gender
interaction, F(1, 60) ¼7.60, p< .01, Z2
p¼.11. Women felt less
1
3
5
menwomen
stereotypical
nonstereotypical
Success beliefs in CS
(1 to 7 scale)
Figure 2. Women’s and men’s beliefs about whether they will be suc-
cessful in computer science (CS) after interacting with a stereotypical
or nonstereotypical computer science role model avatar in Study 2.
Cheryan et al. 5
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
similar to stereotypical (M¼2.38, SD ¼1.24) than nonstereo-
typical role models (M¼4.05, SD ¼1.13), F(1, 60) ¼20.02,
p< .001, Z2
p¼.25, while men felt equally similar to stereo-
typical (M¼3.64, SD ¼1.28) and nonstereotypical (M¼3.71,
SD ¼.91) role models, F(1, 60) < 1, ns. There were no other
significant main effects or interactions.
5
Women’s, but Not Men’s, Success Beliefs Mediated by
Perceived Dissimilarity
Using the moderated mediation macro developed by Preacher,
Rucker, and Hayes (2007) with 5000 bootstrapping resamples,
we found significant Stereotypicality Participant Gender
interactions on perceived similarity and success beliefs (see
above). When controlling for similarity, the Stereotypicality
Participant Gender interaction no longer predicted success
beliefs, b¼.50, SE ¼.31, ns. Conditional indirect effects at
different values of participant gender revealed that for women,
perceived dissimilarity was a significant mediator of the
relationship between stereotypicality and success beliefs, Z¼
2.28, p< .05, whereas for men, this mediation was not signif-
icant, Z¼.60, ns. Women, but not men, contrasted their suc-
cess beliefs away from stereotypical role models, whom they
perceived as dissimilar from themselves.
General Discussion
Interacting with one member of a field, even briefly, can shape
students’ beliefs about their potential for success in that field.
In two studies, we found that STEM role models who projected
stereotypes of the field interfered with women’s beliefs that
they would be successful in STEM fields. Women are routinely
exposed to these stereotypes in the media (e.g., CBS’s popular
The Big Bang Theory, currently in its fourth season), in adver-
tisements (e.g., PEMCO Insurance’s Ponytailed Software
Geek), and even on websites designed to encourage high school
girls to pursue engineering (e.g., http://EngineerYourLife.com
features a female engineer who designs Star Wars video games
and started programming at age 11). This is unfortunate
because many who have achieved success in STEM do not fit
these stereotypes (Borg, 1999), yet the proliferation of such
stereotypical images in society may be preventing the next gen-
eration of potential female scientists from believing they can
achieve success in STEM.
Including a baseline condition in Study 1 revealed that it
was the stereotypical role model who drove the effects. Inter-
estingly, exposure to nonstereotypical STEM role models did
not improve women’s beliefs about their potential for success
over baseline. One explanation may be that nonstereotypical
role models were not relevant enough to our participants
because they were in a different field from their own (Lock-
wood & Kunda, 1997). Another possibility is that we mod-
eled role models on ‘‘the average college student’’ rather
than making them uniquely similar to our participants (e.g.,
same birthday; see Brown et al., 1992; Mussweiler, 2003;
Stapel & Marx, 2007). In both studies, women’s perceived
similarity to the nonstereotypical role model was only at the
midpoint. As a result, they might not have felt similar enough
to the nonstereotypical role models to be influenced by them.
One implication of this work is that selecting ‘‘average stu-
dents’’ as representatives may not change the beliefs of those
we hope to recruit, even if those average students share the
same gender as the potential recruits. Future research should
investigate whether STEM representatives who embody
STEM-counterstereotypic (e.g., feminine) characteristics
improve women’s success beliefs.
Across both studies, female role models were no more
effective in increasing women’s beliefs about their potential
for success than male role models. Role models typically pro-
vide more information about themselves than gender (e.g.,
interests, background), and such individuating information
can override social category information in shaping infer-
ences (Deaux & Lewis, 1984; Eagly & Wood, 1982; Jussim,
Coleman, & Lerch, 1987; Jussim, McCauley, & Lee, 1995;
Krueger & Rothbart, 1988; Locksley, Borgida, Brekke, &
Hepburn, 1980; Rokeach & Mezei, 1966). Even a single
piece of diagnostic information can be enough to prevent
gendered inferences (Eagly & Wood, 1982; see also Kunda
& Oleson, 1995; Locksley et al., 1980). The fact that women
were equally influenced by male and female role models is
also consistent with previous evidence that gender beliefs
need to be salient in order to influence outcomes (Correll,
2004; Deaux & Major, 1987; Shih, Pittinsky, & Ambady,
1999). This may explain why studies have found that male
and female role models are equally effective in inspiring
women and girls to enter STEM fields (Baruch & Nagy,
1977; Canes & Rosen, 1995; de Cohen & Deterding,
2009; Downing, Crosby, & Blake-Beard, 2005; Lunneborg,
1982; Martin & Marsh, 2005) and why work in develop-
mental psychology (e.g., Bobo doll study) demonstrates that
children are equally likely to mimic male and female role
models (see Maccoby & Jacklin, 1974 for a review).
A lack of gender influence in our studies, however,
still resulted in gendered outcomes. Role models in
STEM—whether male or female—who embodied stereo-
types that are incongruent with the female gender role under-
mined women’s beliefs about their ability to be successful in
STEM while leaving men’s beliefs intact. Note that we are
not arguing that women never make better role models. For
women who have already chosen the domain or are otherwise
highly identified with it, female role models improve
women’s attitudes toward STEM (Stout et al., 2011) and pro-
tect their performance when negative stereotypes are salient
(Marx & Roman, 2002; Marx et al., 2005; McIntyre, Paulson,
& Lord, 2003). However, when it comes to recruiting women
into STEM, role model gender may make less of a difference
than whether role models fit stereotypes that are incompati-
ble with the female gender role. Understanding when gender
of role models matters and when STEM stereotypes have
more of an influence will ensure that we are ‘‘rendering
onto the right students the right intervention’’ (Steele,
1997, p. 624).
6Social Psychological and Personality Science 000(00)
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support
for the research, authorship, and/or publication of this article: Work
supported by an NSF CAREER award (DRL-0845110) and a UW
Royalty Research Grant awarded to S. Cheryan.
Notes
1. We did not include drinking, listed as the third most common
hobby for college students. In Study 1, male but not female non-
stereotypical confederates stated sports as one of their hobbies;
in Study 2, both male and female nonstereotypical confederates
stated sports as a hobby.
2. At our university, it is well-known that becoming a computer sci-
ence major requires a track record of success in computer science
classes and competitive admission to the department. As a result,
role models were rated as highly intelligent (M¼6.16, SD ¼.68),
and a 2 (Stereotypicality) 2 (Role Model Gender) ANOVA
on ratings of intelligence revealed no main effects or interactions,
F(1, 53)s < 1, ns.
3. A 2 (Stereotypicality) 2 (Participant Gender) 2 (Role Model
Gender) chi square analysis on those eliminated revealed no signif-
icant effects, all w
2
(1)s < 1, ns. There was therefore no difference in
attrition rates between conditions.
4. Participants were also asked to indicate how attractive they found
their partner on a scale from 1 (not attractive at all)to7(very
attractive) to investigate a potential alternative explanation that
differences in beliefs about success among women were driven
by a desire to distance oneself from unattractive role models. Using
Preacher and Hayes’ (2004) bootstrapping macro to test mediation
with 5000 resamples showed that ratings of attractiveness did not
mediate the relationship between role model stereotypicality and
success beliefs for women, Sobel Z¼1.35, ns.
5. There was a marginal three-way interaction of Stereotypicality
Participant Gender Role Model Gender, F(1, 60) ¼2.96,
p¼.09, Z2
p¼.05, which appeared to be driven by men’s ratings.
When role models were male, men tended to report more similarity
to nonstereotypical than stereotypical role models, p¼.19. However,
when role models were female, men tended to report more similarity
to stereotypical than nonstereotypical role models, p¼.21.
References
Bandura, A. (1997). Self-efficacy: The exercise of control. New York,
NY: W.H. Freeman.
Barbercheck, M. (2001). Mixed messages: Men and women in adver-
tisements in science. In Women, science, and technology: A reader
in feminist science studies (pp. 117-131). London: Routledge.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator
variable distinction in social psychological research: Conceptual,
strategic, and statistical considerations. Journal of Personality and
Social Psychology,51, 1173–1182.
Baruch, R., & Nagy, J. (1977). Females and males in the potential
scientist pool: A study of the early college years. Washington,
DC: National Science Foundation.
Betz, N. E., & Hackett, G. (1981). The relationship of career-
related self-efficacy expectations to perceived career options
in college women and men. Journal of Counseling Psychology,
28,399–410.
Borg, A. (1999). What draws women to and keeps women in
computing? The Annuals of the New York Academy of Sciences,
869, 102–105.
Brown, J. D., Novick, N. J., Lord, K. A., & Richards, J. M. (1992).
When Gulliver travels: Social context, psychological closeness,
and self-appraisals. Journal of Personality and Social Psychology,
62, 717–727.
Buunk, A. P., Peiro´ , J. M., & Griffioen, C. (2007). A positive role
model may stimulate career oriented behavior. Journal of Applied
Social Psychology,37, 1489–1500.
Canes, B. J., & Rosen, H. S. (1995). Following in her footsteps?
Faculty gender composition and women’s choices of college
majors. Industrial and Labor Relations Review,48, 486–504.
Ceci, S. J., Williams, W. M., & Barnett, S. M. (2009). Women’s
underrepresentation in science: Sociocultural and biological con-
siderations. Psychological Bulletin,135, 218–261.
Cejka, M. A., & Eagly, A. H. (1999). Gender-stereotypic images of
occupations correspond to the sex segregation of employment.
Personality and Social Psychology Bulletin,25, 413–423.
Cheryan, S., & Plaut, V. C. (2010). Explaining underrepresentation:
A theory of precluded interest. Sex Roles,63, 475–488.
Cheryan, S., Plaut, V. C., Davies, P. G., & Steele, C. M. (2009). Ambi-
ent belonging: How stereotypical cues impact gender participation
in computer science. Journal of Personality and Social Psychol-
ogy,97, 1045–1060.
Cohen, G., & Garcia, J. (2005). ‘‘I am us’’: Negative stereotypes as
collective threats. Journal of Personality and Social Psychology,
89, 566–582.
Correll, S. J. (2001). Gender and the career choice process: The role of
biased self-assessments. American Journal of Sociology,106,
1691–1730.
Correll, S. J. (2004). Constraints into preferences: Gender, status, and
emerging career aspirations. American Sociological Review,69,
93–113.
Dasgupta, N., & Asgari, S. (2004). Seeing is believing: Exposure to
counterstereotypic women leaders and its effect on the malleability
of automatic gender stereotyping. Journal of Experimental Social
Psychology,40, 642–658.
Davies, P. G., Spencer, S. J., Quinn, D. M., & Gerhardstein, R.
(2002). Consuming images: How television commercials that eli-
cit stereotype threat can restrain women academically and pro-
fessionally. Personality and Social Psychology Bulletin,28,
1615–1628.
de Cohen, C. C., & Deterding, N. (2009). Widening the net: National
estimates of gender disparities in engineering. Journal of Engi-
neering Education,98, 211–226.
Deaux, K., & Lewis, L. L. (1984). Structure of gender stereotypes:
Interrelationships among components and gender label. Journal
of Personality and Social Psychology,46, 991–1004.
Cheryan et al. 7
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
Deaux, K., & Major, B. (1987). Putting gender into context: An inter-
active model of gender-related behavior. Psychological Review,
94, 369–389.
Diekman, A. B., Brown, E., Johnston, A., & Clark, E. (2010). Seek-
ing congruity between goals and roles: A new look at why
women opt out of STEM careers. Psychological Science,21,
1051–1057.
Dotsch, R., & Wigboldus, D. H. J. (2008). Virtual prejudice. Journal
of Experimental Social Psychology,44, 1194–1198.
Downing, R. A., Crosby, F. J., & Blake-Beard, S. (2005). The perceived
importance of developmental relationships on women under-
graduates’ pursuit of science. Psychology of Women Quarterly,
29, 419–426.
Eagly, A. H. (1987). Sex differences in social behavior: A social-role
interpretation. Hillsdale, NJ and England: Lawrence Erlbaum.
Eagly, A. H., & Steffen, V. J. (1984). Gender stereotypes stem from
the distribution of women and men into social roles. Journal of
Personality and Social Psychology,46, 735–754.
Eagly, A. H., & Wood, W. (1982). Inferred sex differences in status as
a determinant of gender stereotypes about social influence. Journal
of Personality and Social Psychology,43, 915–928.
Eastwick,P.W.,&Gardner,W.L.(2009).Isitagame?Evidence
for social influence in the virtual world. Social Influence,4,
18–32.
Ehrlinger, J., & Dunning, D. (2003). How chronic self-views influence
(and potentially mislead) estimates of performance. Journal of
Personality and Social Psychology,84, 5–17.
Fagot, B. J. (1977). Consequences of cross-gender behavior in pre-
school children. Child Development,48, 902–907.
Hill, C., Corbett, C., & St. Rose, A. (2010). Why so few? Women in
science, technology, engineering, and mathematics. Washington,
DC: AAUW.
Inzlicht, M., & Ben-Zeev, T. (2000). A threatening intellectual envi-
ronment: Why females are susceptible to experiencing problem-
solving deficits in the presence of males. Psychological Science,
11, 365–371.
Jussim, L. J., Coleman, L. M., & Lerch, L. (1987). The nature of
stereotypes: A comparison and integration of three theories. Jour-
nal of Personality and Social Psychology,52, 536–546.
Jussim, L. J., McCauley, C. R., & Lee, Y. T. (1995). Why study stereo-
type accuracy and inaccuracy. In Y. T. Lee, L. J. Jussim, &
C. R. McCauley (Eds.), Stereotype accuracy: Toward appreciating
group differences (pp. 3–28). Washington, DC: American Psycho-
logical Association.
Krueger, J., & Rothbart, M. (1988). Use of categorical and individuat-
ing information in making inferences about personality. Journal of
Personality and Social Psychology,55, 187–195.
Kunda, Z., & Oleson, K. C. (1995). Maintaining stereotypes in the
face of disconfirmation: Constructing grounds for subtyping devi-
ants. Journal of Personality and Social Psychology,68, 565–579.
Locksley, A., Borgida, E., Brekke, N., & Hepburn, C. (1980). Sex
stereotypes and social judgment. Journal of Personality and Social
Psychology,39, 821–831.
Lockwood, P. (2006). ‘‘Someone like me can be successful’’: Do col-
lege students need same-gender role models? Psychology of
Women Quarterly,30, 36–46.
Lockwood, P., & Kunda, Z. (1997). Superstars and me: Predicting the
impact of role models on the self. Journal of Personality and
Social Psychology,73, 91–103.
Lockwood, P., Marshall, T. C., & Sadler, P. (2005). Promoting success
or preventing failure: Cultural differences in motivation by posi-
tive and negative role models. Personality and Social Psychology
Bulletin,31, 379–392.
Lunneborg, P. W. (1982). Role model influencers of nontraditional
professional women. Journal of Vocational Behavior,20,
276–281.
Maccoby, E. E., & Jacklin, C. N. (1974). The psychology of sex
differences. Stanford, CA: Stanford University Press.
Margolis, J., & Fisher, A. (2002). Unlocking the clubhouse: Women in
computing. Cambridge, MA: MIT Press.
Martin, A., & Marsh, H. (2005). Motivating boys and motivating girls:
Does teacher gender really make a difference? Australian Journal
of Education,49, 320–334.
Marx, D. M., & Roman, J. S. (2002). Female role models: Protecting
women’s math test performance. Personality and Social Psychol-
ogy Bulletin,28, 1183–1193.
Marx,D.M.,Stapel,D.A.,&Muller,D.(2005).Wecandoit:
The interplay of construal orientation and social comparisons
under threat. Journal of Personality and Social Psychology,
88,432–446.
McIntyre, R. B., Paulson, R. M., & Lord, C. G. (2003). Alleviating
women’s mathematics stereotype threat through salience of group
achievements. Journal of Experimental Social Psychology,39,
83–90.
McKenna, K. Y. A., & Bargh, J. A. (1998). Coming out in the age of
the Internet: Identity ‘‘demarginalization’’ through virtual group
participation. Journal of Personality and Social Psychology,75,
681–694.
Meece, J. L., Parsons, J. E., Kaczala, C. M., & Goff, S. B. (1982). Sex
differences in math achievement: Toward a model of academic
choice. Psychological Bulletin,91, 324–348.
Meece, J. L., Wigfield, A., & Eccles, J. S. (1990). Predictors of math
anxiety and its influence on young adolescents’ course enrollment
intentions and performance in mathematics. Journal of Educa-
tional Psychology,82, 60–70.
Miura, I. T. (1987). The relationship of self-efficacy expectations to
computer interest and course enrollment in college. Sex Roles,
16, 303–311.
Murphy, M. C., Steele, C. M., & Gross, J. J. (2007). Signaling threat:
How situational cues affect women in math, science, and engineer-
ing settings. Psychological Science,18, 879–885.
Mussweiler, T. (2001). ‘‘Seek and ye shall find’’: Antecedents of
assimilation and contrast in social comparison. European Journal
of Social Psychology,31, 499–509.
Mussweiler, T. (2003). Comparison processes in social judgment:
Mechanisms and consequences. Psychological Review,110,
472–489.
National Science Foundation. (2009). TABLE C-4. Bachelor’s
degrees, by sex and field: 1997–2006.Women, minorities, and per-
sons with disabilities in science and engineering: 2009, D. o. S. R.
Statistics. Retrieved from http://www.nsf.gov/statistics/wmpd/
tables.cfm
8Social Psychological and Personality Science 000(00)
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
Okita, S. Y., Bailenson, J., & Schwartz, D. L. (2007, August). Mere
belief in social action improves complex learning. Paper presented
at the Proceedings of the 29th Meeting of the Cognitive Science
Society, Nashville, TN.
Othman, M., & Latih, R. (2006). Women in computer science: No
shortage here!Communications of the ACM,49, 111–114.
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for
estimating indirect effects in simple mediation models. Behavior
Research Methods, Instruments, & Computers,36, 717–731.
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing
moderated mediation hypotheses: Theory, methods, and prescrip-
tions. Multivariate Behavioral Research,42, 185–227.
Rokeach, M., & Mezei, L. (1966). Race and shared belief as factors in
social choice. Science,151, 167–172.
Rudman, L. A., & Fairchild, K. (2004). Reactions to counterstereotypic
behavior: The role of backlash in cultural stereotype maintenance.
Journal of Personality and Social Psychology,87, 157–176.
Sax, L. J. (1994). Predicting gender and major-field differences in
mathematical self-concept during college. Journal of Women and
Minorities in Science and Engineering,1, 291–307.
Schmader, T., Johns, M., & Forbes, C. (2008). An integrated process
model of stereotype threat effects on performance. Psychological
Review,115, 336–356.
Schott, G., & Selwyn, N. (2000). Examining the ‘‘male, antisocial’’
stereotype of high computer users. Journal of Educational Com-
puting Research,23, 291–303.
Sekaquaptewa, D., & Thompson, M. (2003). Solo status, stereotype
threat, and performance expectancies: Their effects on women’s per-
formance. Journal of Experimental Social Psychology,39,68–74.
Shih, M., Pittinsky, T. L., & Ambady, N. (1999). Stereotype suscept-
ibility: Identity salience and shifts in quantitative performance.
Psychological Science,10, 80–83.
Stapel, D. A., & Marx, D. M. (2007). Distinctiveness is key:How different
types of self-other similarity moderate social comparison effects.
Personality and Social Psychology Bulletin,33, 439–448.
Steele, C. M. (1997). A threat in the air: How stereotypes shape
intellectual identity and performance. American Psychologist,
52,613–629.
Stout, J. G., Dasgupta, N., Hunsinger, M., & McManus, M. (2011).
STEMing the tide: Using ingroup experts to inoculate women’s
self-concept and professional goals in science, technology, engi-
neering, and mathematics (STEM). Journal of Personality and
Social Psychology,100, 255–270.
Strenta, C., Elliott, R., Adair, R., Matier, M., & Scott, J. (1994).
Choosing and leaving science in highly selective institutions.
Research in Higher Education,35, 513–547.
Varma, R. (2009). Exposure, training, and environment: Women’s
participation in computing education in the United States and
India. Journal of Women and Minorities in Science and Engineer-
ing,15, 205–222.
Walton, G., & Cohen, G. (2007). A question of belonging: Race,
social fit, and achievement. Journal of Personality and Social
Psychology,92, 82–96.
Wigfield, A., & Eccles, J. S. (2000). Expectancy-value theory of
achievement motivation. Contemporary Educational Psychology,
25, 68–81.
Bios
Sapna Cheryan is an assistant professor of psychology at the Univer-
sity of Washington.
John Oliver Siy is a doctoral student in the department of psychology
at the University of Washington.
Marissa Vichayapai graduated with a BA in psychology from the
University of Washington in 2008 and is the lab manager for the
Stereotypes, Identity, and Belonging Lab at the University of
Washington.
Benjamin J. Drury is a doctoral student in the department of psychol-
ogy at the University of Washington.
Saenam Kim graduated with a BS in psychology from the University
of Washington in 2009 and is pursuing an MSW at Columbia
University.
Cheryan et al. 9
at UNIV WASHINGTON LIBRARIES on April 20, 2011spp.sagepub.comDownloaded from
