Losing its expected communal value: how 
stereotype threat undermines women’s 
identity as research scientists
Authors: Jessi L. Smith, Elizabeth R. Brown, Dustin B. 
Thoman, Eric D. Deemer
This is a postprint of an article that originally appeared in Social Psychology of Education 
in April 2015. 
Smith, Jessi L., Elizabeth R. Brown, Dustin B. Thoman, and Eric D. Deemer. "Losing its expected 
communal value: how stereotype threat undermines women's identity as research scientists." 
Social Psychology of Education (April 2015). 
DOI: https://dx.doi.org/10.1007/s11218-015-9296-8
Made available through Montana State University’s ScholarWorks 
scholarworks.montana.edu 
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
12
3 Losing its expected communal value: how stereotype
4 threat undermines women’s identity as research
5 scientists
6 Jessi L. Smith • Elizabeth R. Brown •
7 Dustin B. Thoman • Eric D. Deemer
8 Received: 13 August 2014 /Accepted: 27 January 2015
9  Springer Science+Business Media Dordrecht 2015
10 Abstract The worry or concern over confirming negative gender group stereo-
11 types, called stereotype threat, is one explanation for women’s worldwide under-
12 representation in undergraduate science classes and majors. But how does
13 stereotype threat translate into fewer women motivated for science? In this quan-
14 titative study with a sample from the US, we use Expectancy Value Theory to
15 examine whether and how stereotype threat concerns might influence women’s
16 science identification. To do this, we collected survey data from 388 women en-
17 rolled in introductory physics (male-dominated) and biology (female-dominated)
18 undergraduate laboratory classes at three universities. We examined multiple
19 indirect effect paths through which stereotype threat could be associated with sci-
20 ence identity and the associated future motivation to engage in scientific research. In
21 addition to replicating established expectancy-value theory motivational findings,
22 results support the novel prediction that one route through which stereotype threat
23 negatively impacts women’s science identity is via effects on perceptions about the
24 communal utility value of science. Especially among women in physics who
A1 J. L. Smith (&)
A2 Department of Psychology, Montana State University, P.O. Box 173440, Bozeman, MT 59717,
A3 USA
A4 e-mail: jsismith@montana.edu
A5 E. R. Brown
A6 University of North Florida, 1 UNF Drive, Jacksonville, FL 32224, USA
A7 e-mail: elizabeth.r.brown@unf.edu
A8 D. B. Thoman
A9 California State University Long Beach, 1250 Bellflower Blvd, Long Beach, CA 90840, USA
A10 e-mail: Dustin.Thoman@csulb.edu
A11 E. D. Deemer
A12 Department of Educational Studies, Purdue University, 100 N. University St., West Lafayette,
A13 IN 47907, USA
A14 e-mail: edeemer@purdue.edu
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
Soc Psychol Educ
DOI 10.1007/s11218-015-9296-8
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
25 expressed greater stereotype threat concerns than women in biology, science
26 identification was lower to the extent that stereotype threat reduced how useful
27 science was seen for helping other people and society. Implications for ways to
28 create an inclusive learning context that combats stereotype threat concerns and
29 broadens undergraduate women’s participation in science are discussed.
30
31 Keywords Gender  Science identification  Science education  Expectancy
32 value  Stereotype threat  Motivation
33
34
35 I’ve always felt that in physics you must have total commitment…it’s not a
36 job; it’s my whole life.
37 —Dr. Chien-Shiung Wu
38 First woman elected as president of the American Physical Society, 19759
40 1 Introduction
41 As Dr. Wu, a Chinese-American woman nuclear physicist, illustrates in this opening
42 quotation, ‘‘scientist’’ can be a personal identity; a fundamental aspect of the self-
43 concept (Smith and White 2001; Markus 1977). To the extent that such
44 identification with science (or any field) predicts achievement and perseverance
45 (e.g., Chemers et al. 2011; Woodcock et al. 2012) it is important to understand the
46 ways in which science identity is developed, undermined, or lays latent as a function
47 of particular social educational structures. Dr. Wu broke many gender (and ethnic)
48 barriers, where the zeitgeist was blatant racism and sanctioned sexism in US society
49 (see Hammond 2009).
50 Worldwide, women are still rare among physicists. For example, women receive
51 as few as 19 % of undergraduate physics degrees in the United States (National
52 Science Board 2012) with similar patterns emerging in tertiary and doctoral degrees
53 worldwide (European Commission 2012). Scandinavian countries graduate the
54 fewest percentage of women physics undergraduates whereas Iran and Italy yield
55 higher percentages (Urry 2000). As the International Union of Pure and Applied
56 Physics Working Group on Women in Physics (2000) reports ‘‘despite cultural
57 differences, the overall situation for women in physics in India, Egypt, Brazil, Latin
58 America, Uruguay, Bolivia and Argentina is much the same. Fewer women than
59 men are receiving degrees in physics.’’ Such underrepresentation renders women
60 who are in physics as ‘‘tokens’’; such solo status can trigger worry or concern about
61 fulfilling gender stereotypes (e.g., Sekaquaptewa and Thompson 2003). The current
62 study examines the extent to which women in undergraduate physics laboratory
63 classes experience these gender stereotype concerns and whether and how such
64 concerns contribute to women’s identification with science and the associated
65 motivation to engage in scientific research.
66 Social identities are comprised of ‘‘people’s knowledge of their memberships in
67 social groups and the emotional significance that they attach to these memberships’’
68 (p. 611, Swann and Bosson 2010). When a situation (e.g., a science classroom)
69 signals that an aspect of one’s social identity (e.g., gender) may be viewed or judged
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
70 through the lens of a negative stereotype, this type of social identity threat is termed
71 ‘‘stereotype threat’’ (Steele et al.). Because being outnumbered by men implicitly
72 activates feelings of stereotype threat (e.g., Inzlicht and Ben-Zeev 2000; Smith and
73 White 2002) when women work or learn in male-dominated fields, they often
74 experience feelings of stereotype threat (Schmader 2002; Schmader et al. 2004;
75 Steele et al. 2002a).
76 Nearly 20 years of research on stereotype threat has established that when gender
77 stereotypes are activated in a high stakes testing situation, women’s math and
78 science performance suffers (for a review see Nguyen and Ryan 2008). Yet, even
79 high ability women who perform well are more likely to opt out of science,
80 technology, engineering, and mathematics (STEM) domains at a greater proportion
81 than their male counterparts, indicating a motivational explanation (e.g., Good et al.
82 2012; Jacobs et al. 2005; Seymour and Hewitt 1995; Smith et al. 2013; Stout et al.
83 2011; Xie and Shauman 2003). Performance, skills, and confidence are important to
84 be sure, but scholars and educators must consider more than just ability to
85 understand science identity (McGee et al. 2012). Indeed, the development of a
86 strong identity as a scientist is shaped by a number of additional motivational
87 factors unrelated to performance (Schmader et al. 2004; Shapiro and Williams 2012;
88 Thoman et al. 2013) and we focus here on the impact of stereotype threat on
89 undergraduate women’s perceptions about the values of science that may be
90 associated with women’s overall science identity.
91 1.1 Science identification
92 Developing and maintaining a stable academic domain identity is paramount to
93 positive educational and career outcomes (e.g., Ahlqvist et al. 2013). There are
94 multiple inputs, pathways, and consequences associated with a student’s identifi-
95 cation with an academic domain (for a review see Osborne and Jones 2011), and
96 cultural norms often exert a strong influence (Swann and Bosson 2010). Yet, few
97 studies have examined whether women’s academic domain identification is
98 associated with stereotype threat specifically (as reviewed by Thoman et al.
99 2013), and no studies that we are aware examine how stereotype threat contributes
100 to women’s science identification.
101 A handful of studies have shown a negative relationship between awareness of
102 (and sensitivity to) stereotypes and domain identity (Ahlqvist et al. 2013; Nosek
103 et al. 2002; Karpinski and Hilton 2001). For example, among undergraduate women
104 living in a dormitory for ‘‘women in science and engineering’’ (WISE), awareness
105 of explicit gender stereotypes about STEM was negatively correlated to identifi-
106 cation with science (Ramsey et al. 2013). Moreover, in situations where stereotype
107 threat is likely to be low (e.g., in the presence of counter-stereotypic role models and
108 experts), science identity is stronger (Stout et al. 2011; Young et al. 2013). One of
109 the few studies that examined the connection between stereotype threat specifically
110 and overall domain identification showed greater levels of stereotype threat among
111 Latino undergraduate students was associated with lower levels of science identity
112 which in turn, predicted a decreased desire to pursue a scientific career after college
113 (Woodcock et al. 2012).
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
114 We build on this work examining the relationship between stereotypes and
115 science identity and ask whether and how the degree to which even slight variations
116 in naturally occurring stereotype threat concerns contribute to undergraduate
117 women’s identification with science specifically, and the relationship between
118 science identity and associated future intention to pursue scientific research with
119 faculty. Given the dearth of literature on this specific association, we draw from
120 established motivational theories, specifically expectancy-value theory, to inform
121 our hypotheses.
122 1.2 Expectancy value theory
123 People feeling vulnerable to stereotype threat not only underperform on high stakes
124 exams such as the Graduate Record Exam (see Nguyen and Ryan 2008 for review),
125 but also suffer decrements in confidence, elevated anxiety, and an overall reduction
126 in expectancies for academic success (e.g., Cadinu et al. 2003, 2005; Stangor et al.
127 1998; for a review see Smith 2004). The role of such performance expectancies in
128 women and girls’ motivation for STEM has a long history within the Expectancy-
129 Value Theory literature (e.g., Eccles 1989, 2009; Eccles and Wigfield 2002; see
130 Wang and Degol 2013 for a review). Expectancy-Value Theory is a multiplicative
131 function used to predict performance and motivation, and it is assumed that
132 expectancy and value are positively related to each other (e.g., Eccles and Wigfield
133 2002). According to the theory, individuals choose, persist, and succeed in career
134 fields/educational domains to the extent they believe that they will do well in the
135 field (expectancies) and the field is relatively valuable (value perceptions).
136 Among children, expectancies for success predict achievement performance
137 (e.g., grades) and perceiving relatively high value, in mathematics or sports for
138 example, predicts motivation (e.g., greater course enrollment and involvement, see
139 Eccles and Wigfield 2002 for a review). Over time, expectancies and value become
140 positively related to each other as children develop into adults and begin to see more
141 and more value in tasks they perform well. Results hold true whether those
142 expectancies or values come about directly from close others (e.g., a mother’s
143 beliefs about her daughter’s math ability, Jacobs and Eccles 1992) or broader socio-
144 cultural norms (e.g., stereotypes, Wang and Degol 2013; Eccles 2005).
145 Thus, Expectancy-Value Theory predicts if stereotype threat is associated with a
146 decline in expectancies for success (operationalized as confidence and anxiety),
147 women’s motivation and identification with science should similarly suffer (Eccles
148 2009). As past work has confirmed the deleterious effects of stereotype threat on
149 confidence (e.g., Cadinu et al. 2003) and anxiety (albeit evidence for anxiety is
150 mixed, see Smith 2004 for a review), we expect to replicate these expectancy
151 findings among undergraduate women in physics lab classes, compared to those
152 enrolled in biology, as a function of increased stereotype threat concerns. We also
153 test whether women’s expectancies for success in their science lab classes (and the
154 contribution of each expectancy aspect in particular) is associated with women’s
155 identification with science and future motivation to engage in scientific research.
156 For the most part, the focus of expectancy-value research on women in science
157 has centered predominantly on the ‘‘expectancy’’ side of the theory (and has
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
158 extended into other theories such as social-cognitive career theory, Lent et al. 1996)
159 with an emphasis on polices and practices aimed at supporting students’ feelings of
160 competence and confidence (Eccles and Wigfield 2002). The role of ‘‘value’’ in the
161 Expectancy-Value Theory equation, has received considerably less empirical
162 attention in the literature. Eccles herself has suggested that understanding value,
163 particularly in terms of how it relates to students’ perceptions of their own identities
164 and social roles, is key to understanding motivated action (Eccles et al. 1999).
165 Indeed, when a student says to herself ‘‘I can, but I don’t want to’’ such a choice
166 likely reflects the relative perceived (low) value of the option (Jacobs et al. 2005).
167 To be sure, there are many different types of values that an academic major, a
168 career or task can afford that will motivate students (i.e., attainment value, intrinsic
169 value, utility value, perceived cost; Eccles 2005). We focus here on ‘‘utility value.’’
170 A discipline or task has utility value if a person perceives the task as affording a
171 means for reaching either a short- or long-term goal, for example, helping other
172 people. Students who are guided to self-generate the ‘‘utility’’ of a task or are
173 provided information about a task’s utility during instruction show more interest in
174 the task, especially those who are low in confidence (e.g., Hulleman and
175 Harackiewicz 2009; Hulleman et al. 2010).
176
177
178
179
180
181
182
183
184
185
186
187
Utility value is a broad category; we focus here on two types. communal and 
agentic utility value (Fiske et al. 2007; Diekman et al. 2011). These two types of 
utility value are fundamental values embedded within most cultures (e.g., Bakan 
1966; Fiske et al. 2007; Helgeson 1994; Pohlmann 2001). Communal value is the 
extent to which a task is other-oriented (i.e., involves working with or helping others 
and creates opportunity for intimate bonds) whereas agentic value is the extent to 
which a task provides autonomy, power, prestige or wealth (Pohlmann 2001). 
Communion is embedded within female (and typically western) gender norms and 
agency within male (typically western) gender norms (e.g., Eagly et al. 2000) and as 
a cause or consequence, there is a large robust finding that men prefer working with 
‘‘things’’ and women prefer working with ‘‘people’’ (see the meta analyses by Su 
et al. 2009).
188 Although American male and female children as young as 5 on up through adults
189 all equally value agentic utility (e.g., money), women and girls are more likely to
190 value communal utility (e.g., Morgan et al. 2001; Weisgram et al. 2010; Konrad
191 et al. 2000), and seeing such value in a given field is associated with women’s career
192 interest and motivation (e.g., Diekman et al. 2010, 2011; McGee and Keller 2007).
193 For example, Weisgram et al. (2010) demonstrated that a novel job described as
194 high in communal utility (defined as having altruistic value) was linked to decreased
195 interest for boys and men, but among girls and women, interest in the novel job was
196 just as high if described as affording altruistic utility (communal) or money utility
197 (agentic). Science jobs are generally perceived as low in communal utility value;
198 science is seen as providing few opportunities to connect with and benefit other
199 people; especially compared to other formerly male-dominated disciplines (medical
200 doctor, lawyer) and female-dominated disciplines (e.g., nursing, education)
201 (Diekman et al. 2010). This perception of science as non-communal is problematic
202 and deters undergraduate American women from a science, career (Diekman et al.
203 2010, 2011). In short (mis)perceptions that science professions are low in communal
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
204 values conveys information that is at odds with what many women value in a career
205 (e.g., Good et al. 2012; Morgan et al. 2001; Smith et al. 2013) and such low
206 communal value is likely an impediment to women’s science identity (Wang and
207 Degol 2013).
208 When stereotype threat is triggered, women become vigilant to cues that they do
209 not belong (e.g., Cheryan et al. 2009; Good et al. 2012). As such, we predicted that
210 women may be particularly likely to perceive science as low in communal value
211 utility when experiencing stereotype threat in their science classroom, which would be
212 a novel finding. This prediction is based in research on ‘‘stereotype threat spillover’’
213 (Carr and Steele 2010) such that coping with stereotype threat can lead women to be
214 loss-averse (Carr and Steele 2010), cause women to downplay the importance of their
215 own feminine characteristics (Pronin et al. 2003) result in a deflated sense of women’s
216 communal ‘‘we’’ sense of self (Keller and Sekaquaptewa 2008), and make salient the
217 majority group’s values (in this case, the majority group is white men who do not
218 highly value the communal utility of science, Smith et al. 2014a, b). It is therefore
219 important to ask if certain science classroom contexts (those more likely to trigger
220 stereotype threat) undermine undergraduate women’s ability to see science disciplines
221 as affording communal utility value, and if so, such low communal utility value
222 should negatively influence women’s willingness to identify with science.
223 1.3 Study overview
224 The goal of the study was to examine whether and how stereotype threat impacts
225 women’s science identity using an Expectancy-Value Theory formulation with a
226 specific focus on communal utility value. We also examined the relationship
227 between science identity and women’s intention to engage in future scientific
228 research (versus motivation for science more generally). Understanding women’s
229 intention to pursue and engage in scientific research as part of their future college
230 studies is especially important considering the need to diversify the scientific
231 workforce to improve discovery, innovation, and remain competitive to meet global
232 scientific needs (McGee et al. 2012; STEMconnector 2012; President’s Council of
233 Advisors on Science and Technology 2012). Women are highly underrepresented in
234 science jobs in general, but are even more likely than men to opt out of research
235 intensive science majors and careers (e.g., Martinez et al. 2007) especially in highly
236 male-dominated science research fields such as physics (NSF 2011).
237 We surveyed undergraduate women at three US universities enrolled in either
238 highly male-dominated physics laboratory classes or female-dominated biology
239 laboratory classes. It was predicted that women in physics lab classes would report
240 greater experiences of stereotype threat than women in biology lab classes because
241 the male-dominated nature of the classes should trigger stereotype threat (Inzlicht
242 and Ben-Zeev 2000; Smith and White 2002). Moreover, it was expected that greater
243 gender stereotype threat concerns would reduce expectancies for success (in line
244 with past research findings). Our novel prediction was that greater stereotype threat
245 concerns would be associated with a reduction in the perceived communal utility
246 value of science that would be related to less willingness to identify with science.
247 We further expected that lower levels of science identity would be associated with
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
248 women’s decreased intention to engage in scientific research with faculty. We also
249 explored the contributing role of an alternative type of utility value (agency).
250 Although less central to the study aim, it was possible that stereotype threat would
251 be linked to greater agentic value perceptions. Recall, agentic utility values are
252 characterized by a self-oriented focus and involves seeking new experiences with a
253 focus on power, competition and/or achievement (Pohlmann 2001). Given science is
254 a ‘‘masculine’’ field that is (or at least is stereotyped as) highly competitive and
255 agentic (Diekman et al. 2010; see also Smith et al. 2013) it is possible that women
256 who have greater gender stereotype threat concerns would be more likely to
257 associate agentic values with science. Alternatively, because agency is already
258 central to the culture of science, it is possible that perceptions of the agentic value of
259 science will be unaffected by stereotype threat concerns. We test both possibilities.
260 2 Method
261 2.1 Participants and procedure
262 A total of 388 women enrolled in introductory biology (n = 198) and physics
263 (n = 190) courses (61.6 % Caucasian; ages 18–39; median age = 20.78; 12.4 %
264 Freshman; 27.6 % Sophomores; 24.7 % Juniors; 32.2 % Seniors; 2.5 % ‘‘other’’) at
265 three different universities (located in the South, West Coast, and Mountain West in
266 the United States of America) participated in this survey in exchange for $10 in
267 compensation. More than one-half (61.9 %) of the participants reported majoring in
268 a STEM discipline, 37.4 % reported majoring in some other discipline, and 0.7 %
269 had not yet declared a major. The introductory biology courses we sampled from
270 were populated primarily by women (67.6 %) whereas the introductory physics
271 courses mainly consisted of men (64.3 %). No differences emerged as a function of
272 data collection site, thus this variable is not discussed further.
273 Course rosters containing student email addresses for students enrolled in biology
274 and physics lower division (100 and 200 level) lab courses were obtained from the
275 registrars’ offices. Email messages inviting students to participate in a ‘‘research
276 motivation study’’ to ‘‘gain understanding about perceptions of science lab classes’’
277 were then distributed to students at the midpoint of the academic term. Participants
278 completed an internet-based survey assessing which lab class they were enrolled in,
279 their feelings of stereotype threat, and measures assessing their confidence in
280 science, anxiety about science, beliefs about the degree to which science fulfills
281 communal and agentic utility values, their future research intentions and, most
282 importantly, their identification with science. Measures were counterbalanced.
283 2.2 Measures
284 2.2.1 Stereotype threat
285 Participants rated themselves on three different items adapted from the Impression
286 and Threat Concern Scale (Marx and Goff 2005; see also Keller and Sekaquaptewa
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
287 2008) on scales ranging from 1 (strongly disagree) to 7(strongly agree). The
288 adapted items were as follows: ‘‘I worry that my ability to perform well in my class
289 is affected by my gender;’’ ‘‘I worry that if I perform poorly in my science lab class,
290 others will attribute my poor performance to my gender;’’ and ‘‘I worry that,
291 because I know the negative stereotype about women and science ability, my
292 anxiety about confirming this stereotype will negatively influence how I perform in
293 my science lab class.’’ The three items were summed to form a stereotype threat
294 (ST) index (a = 0.90).
295 2.2.2 Confidence in science
296 As one index of expectancy for success, we assessed science confidence using the
297 confidence in learning science (CLS) subscale of the Science Motivation
298 Questionnaire (SMQ; Glynn and Koballa 2006). Participants respond to five items
299 on Likert scales ranging from 1 (never) to 5 (always). The five confidence items are
300 as follows: (a) ‘‘When I am in a college science course I expect to do as well as or
301 better than other students in the science course’’; (b) ‘‘When I am in a college
302 science course I am confident I will do well on the science labs and projects’’;
303 (c) ‘‘When I am in a college science course I believe I can master the knowledge and
304 skills in the science course’’; (d) ‘‘When I am in a college science course I am
305 confident I will do well on the science tests’’; and (e) ‘‘When I am in a college
306 science course I believe I can earn a grade of ‘A’ in the science course.’’ These
307 items were summed to create a confidence in science index (a = 0.84).
308 2.2.3 Anxiety about science
309
310
311
312
313
314
315
316
317
318
319
320
321
As a second index to expectancy for success, we used the 5-item ‘‘anxiety about
science assessment’’ (ASA) subscale of the SMQ (Glynn and Koballa 2006) t o
measure science anxiety. ASA items include: (a) ‘‘I am nervous about how I will do
on the science tests’’; (b) ‘‘I worry about failing the science tests’’; (c) ‘‘I become
anxious when it is time to take a science test’’; (d) ‘‘I am concerned that other
students are better in science’’; and (e) ‘‘I hate taking science tests’’. These items
were originally designed to be reversed-scored and summed to contribute to an
overall motivation score on the SMQ. However, factor analytic work with the SMQ
has shown ASA items to produce strong factor loadings, with standardized coefficients
ranging from 0.63 to 0.81 (Glynn, Taasoobshirazi, & Brickman 2009). The ASA is 
therefore believed to possess sufficient enough construct validity to be used as a stand-
alone measure. Responses are scored on a Likert scale ranging from 1 (never) t o 5
(always). These items were summed to form an anxiety index (a = 0.83).
322 2.2.4 Perceptions of science utility values
323 Participants answered two questions assessing how much they believed that science
324 careers afforded communal and agentic utility values on scales ranging from 1 (not
325 at all) to 7 (extremely) (taken from Diekman et al. 2010, 2011). The communal item
326 was as follows: ‘‘How much do you believe that science affords working with
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
327 people, helping others, and serving the community?’’ The agentic item was as
328 follows: ‘‘How much do you believe that science affords power, achievement, and
329 seeking new experiences or excitement?’’
330 2.2.5 Future science research intentions
331 Participants rated their intention to engage in science research in the future on three
332 items on scales ranging from 1 (not likely at all) to 5 (very likely) as used in Smith
333 et al. (2014b). The items were as follows: ‘‘How likely would you be to pursue
334 undergraduate research opportunities?’’ How likely would you be to volunteer to
335 work in a faculty research lab?’’ ‘‘How likely would you be to volunteer to work on
336 a faculty member’s research team?’’ ‘‘How likely would you be to volunteer to work
337 on a faculty member’s research?’’ These items were summed to form a research
338 intentions index (a = 0.96).
339 2.2.6 Identification with science
340 Items from a group identification measure developed by Doosje et al. (1995) were
341 adapted to assess the extent of participants’ identification with science. The original
342 4-item measure was developed for the purpose of tapping the cognitive and affective
343 dimensions of identification with an academic domain (i.e., ‘‘I see myself as a
344 psychology student’’). For the purposes of this study, we modified the original
345 measure and added two additional items to provide broader coverage of the above-
346 noted dimensions of identification. Items in the present study were rated on a 1
347 (strongly disagree) to 7 (strongly agree) scale as follows: (a) ‘‘I see myself as a
348 science student’’; (b) ‘‘I am pleased to be a science student’’; (c) ‘‘I feel strong ties
349 with other science students’’; (d) ‘‘I identify with other science students’’; (e) ‘‘I feel
350 that being a science student is an important reflection of who I am’’ (added item
351 from Luhtanen and Crocker 1992; Walsh and Smith 2007); and (f) ‘‘I don’t act like
352 the typical science student’’ (reverse-scored; see Walsh and Smith 2007). These
353 items were summed to form an identification with science index (a = 0.90).
354 3 Results
355 3.1 Analyses overview
356 Table 1 presents the descriptive statistics and correlations for all of the study
357 variables. Two structural equation models were analyzed using Mplus version 7
358 (Muthen and Muthen 2012) using the maximum likelihood estimation model. Model
359 1 provides a test for our conceptual model, with a focus on the relationship between
360 stereotype threat effects and science identity occurring through communal utility
361 value (see Fig. 1). In model 2, we add agentic utility as a secondary value process
362 variable alongside communal utility value, in order to test whether potential effects
363 involving communal utility value hold even when accounting for students’ agentic
364 values (see Fig. 2). Because the models that we present are not nested, we cannot
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
Table 1 Correlations, means, standard deviations, and ranges for model variables
Subscale 1 2 3 4 5 6 7
1 Stereotype threat – -0.25*** 0.25*** -0.15** -0.06 -0.03 -0.12*
2 Confidence – -0.34*** 0.20*** 0.26*** 0.37*** 0.50***
3 Anxiety – -0.06 -0.01 -0.07 -0.14**
4 Communal utility value
perception
– 0.43*** 0.27*** 0.38***
5 Agentic utility value
perception
– 0.31*** 0.38***
6 Future science research
intentions
– 0.56***
7 Science identification –
R2 values
Model 1 R2 values 0.02 0.07 0.06 0.02 – 0.32 0.32
Model 2 R2 values 0.02 0.07 0.06 0.02 0.004 0.31 0.33
Total mean 5.96 19.25 17.37 5.68 5.98 14.09 31.21
Biology mean 5.36 19.64 17.37 5.69 6.06 14.21 31.14
Physics mean 6.58 18.80 17.38 5.67 5.91 13.95 31.28
Total standard deviation 4.25 3.56 4.38 1.13 1.04 4.81 7.67
Biology mean 3.75 3.49 4.13 1.18 1.00 4.97 8.54
Physics mean 4.64 3.60 4.65 1.07 1.08 4.63 6.65
Range 3–21 5–25 5–25 1–7 1–7 5–20 6–42
** p\ 0.01; *** p\ 0.001
Fig. 1 Model 1: Testing perceived communal utility value of science. Path coefficients represent the
statistically significant standardized estimates from the structural equation analysis. Dashed lines indicate
non-significant paths in the model. R2 values for the variables ranged from 0.02 (Stereotype Threat;
Communal Utility Value Perception) to 0.32 (Science Identity; Future Science Research Intentions).
Note: **p\ 0.01; ***p\ 0.001
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
365 statistically compare model fit. Instead we indirectly compare the models based on
366 the fit indices for each model (v2/df, CFI, RMSEA, and SRMR).
367 3.2 Testing the role of perceived communal utility value: model 1
368 Overall, model fit statistics indicate that our conceptual model provides a good fit
369 for the data. Three indices for this model suggest a good fit (CFI = 0.967;
370 RMSEA = 0.062; SMR = 0.04), and one indicator suggests that this is an
371 acceptably fitting model (v2/df = 2.513). Because the overall model achieved
372 good fit, we can explore the specific theoretical predictions through direct and
373 indirect paths between variables, as illustrated in Fig. 1.
374 The first prediction tested in our conceptual model is that women in physics
375 classes would report higher levels of stereotype threat than those in biology classes.
376 As seen in the significant relationship between the two variables at the left side of
377 Fig. 1, this hypothesis was confirmed (means for biology and physics classes are
378 presented in Table 1).
379 The next relationships tested in the model involved the relationship among
380 stereotype threat and women’s expectancy-value motivation variables. Among the
381 expectancy variables, replicating previous findings, greater feelings of stereotype
382 threat significantly predicted lower confidence and higher anxiety. Further, central
383 to the current study’s predictions, greater feelings of stereotype threat also
384 significantly predicted lower perceptions of communal utility value of science
385 among these women science students.
386 Next, the model tests whether the expectancy-value variables were associated
387 with science identity. As seen in Fig. 1, although both confidence and anxiety
388 significantly correlated with science identity in the expected directions (see
389 Table 1), when both variables were accounted for in the model, only confidence,
Fig. 2 Model 2: The additional contribution of perceived agentic utility value of science. Path
coefficients represent the statistically significant standardized estimates from the structural equation
analysis. Dashed lines indicate non-significant paths in the model. R2 values for the variables ranged from
0.02 (Stereotype Threat; Communal Utility Value Perception) to 0.33 (Science Identity). Note: p\ 0.10;
*p\ 0.05; **p\ 0.01; ***p\ 0.001
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
390 and not anxiety, was significantly related to science identity. More importantly for
391 the current study’s predictions, greater perceived science communal utility value
392 was significantly associated with greater science identity.
393 We next examined indirect effects, which can reveal the influence of a variable
394 that a direct effect test might miss (e.g., Preacher and Hayes 2004). Indeed, the tests
395 of model indirect effects also demonstrate that stereotype threat was significantly
396 indirectly associated with science identification through two paths. Consistent with
397 others’ theoretical predictions (e.g., Cadinu et al. 2003), the first significant indirect
398 path suggests that stereotype threat predicts lower science identity through its
399 associated with confidence (bootstrapped, 95 % CI -0.35 to -0.13). In addition, we
400 are the first to predict and provide evidence that even small increases in stereotype
401 threat indirectly relates to a decrease in science identity through its association with
402 perceived communal utility value of science (bootstrapped, 95 % CI -0.14 to -
403 0.02).
404 Lastly, as seen in Fig. 1, the model demonstrates that future intentions to pursue
405 research are directly predicted both by science identity and confidence. As with
406 science identity, tests of indirect effects suggest that stereotype threat significantly
407 indirectly relates to lower research intentions through multiple paths. First, a
408 significant path through both confidence (bootstrapped, 95 % CI -0.087 to -0.014)
409 and confidence and science identification emerged (bootstrapped, 95 % CI -0.107
410 to -0.038). More important for the current study hypotheses, a significant path from
411 stereotype threat to science research intentions emerged through perceived
412 communal utility value and science identity (bootstrapped, 95 % CI -0.042 to -
413 0.006). Thus, although perceived communal utility value did not directly influence
414 research intentions (despite their positive correlation, see Table 1), it did play an
415 indirect role along with science identity in explaining how feelings of stereotype
416 threat predict women’s lower intentions to engage in science research.
417 3.3 Testing the possible contribution of perceived agentic utility value: model 2
418 Although our conceptual model emphasizes the contributions of perceived
419 communal utility value to women’s science identity, agentic values (high pay,
420 opportunities for pursuing one’s curiosity, and intrinsic science interests) are the
421 values that the majority of people consider as centrally important to science
422 motivation, identity, and success (Dyer and McWhinnie 2011; McGee and Keller
423 2007; Smith et al. 2014b; Webb et al. 2002). Thus, it is possible that when
424 accounting for agentic values in our model, the role of communal utility value
425 perceptions could diminish. We test this empirical question by estimating Model 2,
426 which is identical to Model 1 but with the addition of perceived agentic utility value
427 as a parallel value alongside communal utility value (see Fig. 2).
428 As seen in Fig. 2, although greater agentic value perceptions significantly predict
429 greater science identity, agentic value perceptions are not influenced by stereotype
430 threat nor do they play an indirect role in effects of stereotype threat on science
431 identity. More importantly for our theoretical framework, including agentic value
432 perceptions in the model does not change effects of communal utility value
433 perceptions. Both the direct and indirect effects involving communal utility value
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
434 reported in Model 1 remain significant in Model 2. Further, examination of model fit
435 statistics suggests that adding agentic utility value does not improve fit of the
436 empirical model (CFI = 0.944; RMSEA = 0.079; SMR = 0.061; v2/df = 3.412).
437 Each of the model fit indices indicate that Model 1 fits the data better than Model 2.
438 Thus, adding perceived agentic utility value to the empirical model does not
439 diminish the importance of communal utility value.
440 3.4 Testing alternative reverse pathway models
441 Although we argue that stereotype threat concerns are associated with a reduction in
442 expectancies for success and the perceived communal utility value of science, the
443 correlational nature of our data render it important to test reverse pathways among
444 the variables. We did this in two ways. To start, we tested whether the type of class
445 (physics versus biology) was associated with confidence, anxiety, and communal
446 utility value, and if in turn, these variables were associated with concerns about
447 stereotype threat, with stereotype threat concerns being more proximally associated
448 with science identity and motivation. Model fit statistics show that reversing the
449 relationships in this way resulted in poor fit (CFI = 0.57; RMSEA = 0.20;
450 SRMR = 0.14; v2/df = 17.05) suggesting that this pathway was not a significantly
451 supported possibility.
452 Next, we tested whether the type of class (physics versus biology) was associated
453 with communal utility value first, and if in turn this variable was associated with
454 concerns about stereotype threat, with stereotype threat concerns then being
455 associated with expectancies (confidence and anxiety), science identity, and
456 motivation. Again, the model fit statistics show that reversing the model in this
457 way also resulted in a poor fit (CFI = 0.84; RMSEA = 0.12; SMR = 0.09; v2/
458 df = 6.76).
459 Taken together, we argue that although the causal pathways cannot be confirmed
460 with complete certainty because of the correlational nature of the data, testing these
461 alternative pathway models offer some evidence as to the veracity of the direction of
462 effects as shown in Model 1. We return to this topic in the discussion.
463 4 Discussion
464 Understanding factors that influence women’s science identity is important to
465 understanding women’s pursuit and persistence in male dominated STEM fields
466 (e.g., Osborne and Jones 2011; Ramsey et al. 2013). We set out to examine whether
467 and how women’s science identity is associated with feelings of stereotype threat,
468 especially in male-dominated physics classes. Drawing from Expectancy-Value
469 Theory (e.g., Eccles and Wigfield 2002), we focused on the associations among
470 science identity, stereotype threat, and women’s perceptions about the communal
471 utility value of science –perceptions about how much science affords the chance to
472 work with and help other people. As predicted, our findings showed that in line with
473 past experimental work on stereotype threat (e.g., Murphy et al. 2007) feelings of
474 stereotype threat were relatively greater in the (male-dominated) physics lab classes
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
475 versus the (female-dominated) biology lab classes. This is important, as there are
476 clearly essential differences within various types of science disciplines that should
477 not be overlooked, which can happen when ‘‘science’’ fields are merged into one
478 broad category. The relationship between science identity and stereotype threat was
479 significantly associated with beliefs about how much science fulfills communal
480 utility value. The addition of perceived agentic utility value did not add to or change
481 the pattern of results. Science identification was also found to be important in
482 predicting future intentions to engage in science research, which was generally
483 lower among women who reported feeling stereotype threat.
484 Our findings are among only a handful to date that examine women’s science
485 identity as an outcome predicted by stereotype threat, and are one of the first to
486 demonstrate one of the ways in which even small differences in stereotype threat
487 experiences in a naturalistic setting contribute to women’s science identity.
488 Women’s science identity was associated indirectly with stereotype threat via
489 perceptions in how much science can and does involve working with and helping
490 other people. When women reported fewer stereotype threat concerns, they were
491 more likely to see that science has communal utility value, and in turn greater
492 perceived communal utility value was associated with elevated science identity. In
493 contrast, when women reported greater stereotype threat concerns, which was more
494 often the case in male-dominated physics classes compared to biology classes, they
495 also viewed science as affording fewer communal utility opportunities; lower
496 communal utility value was associated with lower science identity. These findings,
497 we hope, have both theoretical and applied implications.
498 Theoretically, our model merged together three different literatures to test
499 predictions about influences on undergraduate women’s science identity: stereotype
500 threat (e.g., Steele 1997), Expectancy-Value Theory (e.g., Eccles and Wigfield
501 2002), and research on communal goals and interpersonal values (e.g., Diekman
502 et al. 2011; Maddux and Brewer 2005). Specifically, our findings contribute to the
503 literature on what is known about the ways in which stereotype threat, women’s
504 science identification, and motivation are related. What we do know from the
505 relatively little existing past research is that stereotype threat reduces women’s
506 motivation for a stereotype-relevant domain. To be sure, motivation is related to
507 domain identification (e.g., Smith and White 2001). Much of this evidence however
508 is indirect; for example, when considering an upcoming engineering conference,
509 women who see that the overwhelming majority of conference attendees are men
510 (versus gender equal) report less motivation to participate in the conference
511 (Murphy et al. 2007). Indeed, women’s motivation to consider a novel science field
512 of study declines when that field is presented as male-dominated versus gender
513 equal (Smith et al. 2013). Even when women are not confronted directly with their
514 token status, just sitting in a classroom that is decorated with male-stereotypic items
515 activates stereotype threat and results in women reporting less interest in pursuing
516 (computer) science (Cheryan et al. 2009). These studies assume, but do not measure
517 or manipulate, stereotype threat. Actual direct assessment of stereotype threat and
518 motivation is relatively rare (cf., Davies, Spencer, Quinn, and Gerhardstein 2002;
519 Schmader et al. 2004; Thoman et al. 2008) and although domain identification and
520 motivation are typically positively related (Osborne and Walker 2006), rarer still is
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
521 the inclusion of domain identification as the specific variable under study. Thus, one
522 contribution of the current study was determining that undergraduate women in US
523 physics laboratory classes (which were male-dominated) experience stereotype
524 threat to a greater extent than women in more gender equal or female-dominated
525 science classes (e.g., biology); such stereotype threat concerns were associated with
526 women’s depleted identification with science.
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
It is well documented within the stereotype threat literature that domain
identification moderates the effect of stereotype threat on performance (e.g.,
Aronson et al. 1999; Cadinu et al. 2003) and as such, domain identification is
typically used as a participant selection criterion (e.g., Spencer et al. 1999), as a
predictor variable (e.g., Forbes et al. 2008; Lawrence and Crocker 2009; Osborne
and Walker 2006), or as a covariate (e.g., Smith and White 2002) to isolate the
effects of stereotype threat on other outcomes. Little research exists on the
stereotype threat-identification-motivation relationship (see Thoman et al. for
review). Understanding that stereotype threat decreases interest in pursuing or
persisting in science is important, as measures of interest are indicative of actual
degree selection, degree completion, and career pursuit (for a review see Nye et al.
2012). Yet, for women, science is a ‘‘leaky pipeline’’ (Blickenstaff 2005; Wickware
1997) whereby even momentary gains in interest or career intentions do not ensure 
persevering in science. What is also needed is for women to develop strong feelings
of a personal connection to science, whereby much of her sense of self is defined by
her participation in science. Our results suggest that one important pathway through
which science identification was influenced by stereotype threat was via expectan-
cy-value differences, in particular in perceptions about the communal utility value
of science.
546 Indeed, our findings extend past work on Expectancy-Value Theory by focusing
547 on the association between stereotype threat concerns and a specific type of (gender
548 conflated) utility value: communal values. Women as a group are generally more
549 likely to perform better and enjoy working on tasks that are perceived as
550 purposefully serving society and involve other people (see Su et al. 2009 for a
551 review). Yet, science is often stereotyped as not affording such communal goals
552 (e.g., Diekman et al. 2010, 2011). Add to this that stereotype threat triggers
553 vigilance to cues that women do not belong in male-dominated domains (e.g.,
554 Kaiser et al. 2006; Murphy et al. 2007; Steele et al. 2002b). Putting this all together,
555 we predicted and found that this unique merging of literatures was useful for
556 informing contributing factors to women’s science identity. Certainly it is possible
557 that our combined theoretical framework is only applicable to science identity in
558 women. Moreover, we only examined two different types of utility value and there
559 are many other values worthy of investigation. Our hope is that other domain
560 identities and other types of values might also be served by considering these
561 theoretical perspectives together.
562 Our findings also have practical implications for designing empirically based
563 interventions aimed at reducing stereotype threat and enhancing women’s science
564 identity (e.g., Smith and Hung 2008). In line with recent work showing that helping
565 parents see the personal utility of math and science improves their high-school sons
566 and daughters’ motivation for the fields (Harackiewicz et al. 2012) our findings go
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
567 one step further and suggest that not all ‘‘utility’’ values are equal when considering
568 improving women’s science identity. Instead, communal utility in particular
569 (compared to agentic utility for example) may be important for educators and
570 scholars to consider when designing and delivering classroom interventions aimed
571 at broadening the participation of women in science (and other underrepresented
572 minority groups, see Smith et al. 2014a). Our results imply that women’s
573 identification with science may be enhanced when the communal utility value of
574 physics is made salient, which remains to be tested in future research.
575 Additionally, our focus on the relationship between science identity and science
576 research intentions adds to the emerging literature on broadening women’s
577 participation as scientific and biomedical researchers. In the European Union—
578 27, as few as 33 % of all science researchers in 2009 were women (European
579 Commission 2012). And for those women who are science researchers, there
580 remains a worldwide gender disparity in access to resources (Ivie and Tesfaye
581 2012). For example, women science researchers in the US made up only 25 % of
582 NIH and 23 % of NSF 2007 grant awards (Nature Neuroscience 2010). Such
583 underrepresentation has clear implications not only for social justice and equal
584 access but also for workforce development, economics, discovery, and innovation
585 (e.g., McKinsey and Company 2007). Individuals might embark on the pursuit of
586 careers in scientific research because they are inherently interested in developing
587 their understanding of the laws and phenomena that govern the operation of the
588 natural world. However, such reasons for conducting research often change
589 depending on the people and situations they encounter, leading to more complicated
590 patterns of motivation for conducting research (Smith et al. 2014b). Undergraduate
591 research experience is one way science interest can be captured and sustained
592 especially among underrepresented students (Hurtado et al. 2009). Our findings
593 point to science identity as a pivotal place in the process that may help understand
594 women’s research intentions in lower level science classes: research motivation
595 appears to be channeled through science identity development and the perception
596 that science affords fulfillment of important communal values.
597 4.1 Limitations and future directions
598 Our research focused specifically on women’s science identity, and not men’s,
599 largely because of women’s underrepresented status in science (National Science
600 Board 2012; European Commission 2012) and because the gender stereotypes
601 regarding science are negative towards women as a group (e.g., Rahm and
602 Charbonneau 1997; see also Nosek et al. 2009). Nevertheless, the world is in dire
603 need of more qualified people participating in science in general and thus it is
604 important for future work to examine how positive gender stereotypes about men,
605 for example, may ‘‘boost’’ versus trigger a ‘‘choking under pressure’’ reaction
606 among majority men that contributes to men’s science identity (e.g., Smith and
607 Johnson 2006). It is also the case that some cultures highly value and uphold
608 communal ideals for both men and women, and to the extent that ethnic minority
609 male students (e.g., US Native American men) identify with those cultures they too
610 may experience similar science identity processes as our findings here (Smith et al.
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
611 2014a). Research on the intersectionality of race, gender, and social class based
612 stereotype threat, is key to advancing our understanding of science identity (e.g.,
613 Tine and Gotlieb 2013).
614 Another important limitation is the correlational nature of the data. Although our
615 conceptual model makes causal predictions, conclusions about causality, direct
616 effects, or the direction of causality should not be inferred from correlational data.
617 One benefit of using structural equation modeling is that associations between
618 variables can be decomposed through multiple a priori specified pathways, allowing
619 for comparisons of multiple indirect paths. This strength was evident in the current
620 findings as, for example, both confidence and anxiety were correlated with science
621 identity but only confidence (and not anxiety) served a significant indirect effect
622 role in the association between stereotype threat and science identity. In addition,
623 although agentic and communal utility value were correlated with each other and
624 both correlated with science identity, only communal (and not agentic) utility value
625 played a role in the relationship between stereotype threat and science identity.
626 What is more, we tested two possible alternative models that reversed the direction
627 of effects, and found no support for those models. Thus, although causality and
628 directionality cannot be inferred, our study examined the multiple pathways of
629 correlational decompositions for the variables surveyed.
630 The stereotype threat literature is rich with excellent experiments establishing its
631 effects and correlates in controlled environments (e.g., Aronson et al. 1999; Croizet
632 et al. 2001; Derks et al. 2008); for example, being outnumbered in a stereotype
633 threatening situation decreases feelings of belonging and connection (e.g., Murphy
634 et al. 2007) and is associated with feelings of a hostile work and learning
635 environment (e.g., Oswald and Harvey 2000. Our examination of multiple paths
636 from stereotype threat to science identity in a naturalistic setting we think
637 complement and extend the existing literature. The next step for research is to
638 consider the overlap and distinctions among the experimentally determined
639 mechanisms involved with stereotype threat effects (e.g., Smith 2004) and
640 communal utility value beliefs. In this way, we can learn more about ‘‘third
641 variable’’ effects that were untested in the current study. For example, perhaps
642 biology classes are more likely than physics classes to have women faculty as
643 instructors. If so, a stereotype inoculation perspective (Dasgupta 2011) would
644 predict that female role models within biology (versus physics), for example, might
645 buttress associated reductions in how much female students think science involves
646 working with and helping other people; this is a key empirical question for future
647 research.
648 Our project findings are necessarily limited by the measurement instruments we
649 selected, for better or worse. Although we were careful to select measures
650 established in the literature, it is true that self-reported explicit measures of science
651 identity might not always capture the level of importance and value of the domain to
652 the overall self (Osborne and Jones 2011). Some scholars instead have begun to
653 focus on implicit science identity as a more holistic and predictive measure of
654 science outcomes (e.g., Nosek et al. 2002; Ramsey et al. 2013). Future research
655 would benefit from continuing to explore the nuanced differences in implicit versus
656 explicit science identity and the relationship between the two. For example, which
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
657 type of identity is necessary (albeit not sufficient) for the development of the other?
658 Which type of science identity, or combination of the two, might buffer, versus
659 make a group member vulnerable, to stereotype threat effects (e.g., Aronson et al.
660 1999; Osborne and Jones 2011)? Indeed, we approached science identity from a
661 stereotype threat framework, which has pros and cons. There is a need for empirical
662 data on the link between stereotype threat and science identity, as we reviewed
663 previously—to date most evidence for this link is indirect. Yet it is also true that
664 naturalistic data might not shine a bright light on this link. People may not be
665 especially willing to admit to or even have insight into personal experiences with
666 prejudice and stereotypes (Crosby 1984) and as such, overall mean scale levels
667 might be low in field studies of science identity and stereotype threat. This was
668 indeed the case in our own study. Yet, our findings suggest that even small
669 differences in stereotype threat concerns have meaningful associations with science
670 identification, and modest differences in science identity are associated with
671 important differences in future research intentions.
672 What is the ideal level of science identity for women working in male-dominated
673 science fields such as physics? This is an empirical question. As Osborne and Jones
674 (2011) point out, for example, the very type of self-identity educators are interested
675 in fostering is often the very same identity that makes someone vulnerable to
676 stereotype threat effects (Aronson et al. 1999; Schmader 2002; Steele 1997). The
677 discovery of some optimal amount of science identity among underrepresented
678 students is a critical question to consider in future research. The answer likely
679 involves a motivational experiences feedback loop (Thoman et al. 2013) whereby
680 science identity becomes tied to belonging, interest, and activity engagement in
681 ways that stereotype threat concerns both undermine the experience (via perfor-
682 mance avoidance goals) and enhance the experience (by cueing an unsafe and
683 adverse negative environment that should be exited).
684 5 Conclusion
685 As was illustrated by Dr. Wu at the outset of our paper, for some science can define
686 one’s life, and this identification with science can drive persistence and commitment
687 to the field. Our study is among the first to examine how stereotype threat concerns
688 shape science identification and how science identification is associated with the
689 motivation to engage in scientific research for an underrepresented group of people.
690 Combining three different literatures (Stereotype Threat, Expectancy-Value, and
691 Communal Goals and Values), this study provides evidence that women’s science
692 identity is shaped by their experience of stereotype threat, with the relationship
693 between stereotype threat and science identification being explained by the
694 perceived communal utility value of science. Implications for our results are
695 twofold: (1) the importance of removing stereotype threat cues and triggers and (2)
696 emphasizing the ways in which science contributes to and helps others as a means to
697 enhancing women’s identification with science. Both implications are in the service
698 of ultimately broadening the participation of women in male-dominated science
699 research fields.
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
700 Acknowledgments We are grateful to Justin Chase and other members of the Motivation and Diversity
701 Lab for assistance with participant recruitment. This project was supported by a grant (award number
702 HRD-1036767) from the Gender in Science and Engineering Division of the National Science
703 Foundation. Any opinions, findings, and conclusions or recommendations expressed in this material are
704 our own and do not necessarily reflect the views of the National Science Foundation.
705
706 References
707 Ahlqvist, S., London, B., & Rosenthal, L. (2013). Unstable identity compatibility: How gender rejection
708 sensitivity undermines success of women in science, technology, engineering, and mathematics
709 fields. Psychological Science, 24, 1644–1652. doi:10.1177/0956797613476048.
710 Aronson, J., Lustina, M. J., Good, C., & Keough, K. (1999). When White men can’t do math: Necessary
711 and sufficient factors in stereotype threat. Journal of Experimental Social Psychology, 35, 29–46.
712 doi:10.1006/jesp.1998.1371.
713 Bakan, D. (1966). The duality of human existence: An essay on psychology and religion. Chicago: Rand
714 McNally.
715 Blickenstaff, J. C. (2005). Women and science careers: Leaky pipeline or gender filter? Gender and
716 Education, 17, 369–386. doi:10.1080/09540250500145072.
717 Cadinu, M., Maass, A., Frigerio, S., Impagliazzo, L., & Lationotti, S. (2003). Stereotype threat: The effect of
718 expectancy on performance. European Journal of Social Psychology, 33, 267–285. doi:10.1002/ejsp.145.
719 Cadinu, M., Maass, A., Rosbianca, A., & Kiesner, J. (2005). Why do women underperform under
720 stereotype threat? Evidence for the role of negative thinking. Psychological Science, 16, 572–578.
721 doi:10.1111/j.0956-7976.2005.01577.x.
722 Carr, P. B., & Steele, C. M. (2010). Stereotype threat affects financial decision making. Psychological
723 Science, 21, 1411–1416. doi:10.1177/0956797610384146.
724 Chemers, M. M., Zurbriggen, E. L., Syed, M., Goza, B. K., & Bearman, S. (2011). The role of efficacy
725 and identity in science career commitment among underrepresented minority students. Journal of
726 Social Issues, 67, 469–491. doi:10.1111/j.1540-4560.2011.01710.x.
727 Cheryan, S., Plaut, V. C., Davies, P. G., & Steele, C. M. (2009). Ambient belonging: How stereotypical
728 cues impact gender participation in computer science. Journal of Personality and Social
729 Psychology, 97, 1045–1060. doi:10.1037/a0016239.
730 Croizet, J. C., De´sert, M., Dutre´vis, M., & Leyens, J. (2001). Stereotype threat, social class, gender, and
731 academic under-achievement: When our reputation catches up to us and takes over. Social
732 Psychology of Education, 3–4, 295–310. doi:10.1023/A:1011336821053.
733 Crosby, F. (1984). The denial of personal discrimination. American Behavioral Scientist, 27, 371–386.
734 doi:10.1177/000276484027003008.
735 Dasgupta, N. (2011). Ingroup experts and peers as social vaccines who inoculate the self-concept: The
736 stereotype inoculation model. Psychological Inquiry, An International Journal for the Advancement
737 of Psychological Theory, 22(4), 231–246. doi:10.1080/1047840X.2011.607313.
738 Derks, B., Inzlicht, M., & Kang, S. (2008). The neuroscience of stigma and stereotype threat. Group
739 Processes & Intergroup Relations, 11, 163–181. doi:10.1177/1368430207088036.
740 Diekman, A. B., Brown, E. R., Johnston, A. M., & Clark, E. K. (2010). Seeking congruity between goals
741 and roles: A new look at why women opt out of science, technology, engineering, and mathematics
742 careers. Psychological Science, 21, 1051–1057. doi:10.1177/0956797610377342.
743 Diekman, A. B., Clark, E. K., Johnston, A. M., Brown, E. R., & Steinberg, M. (2011). Malleability in
744 communal goals and beliefs influence attraction to STEM careers: Evidence for a goal congruity
745 perspective. Journal of Personality and Social Psychology, 101, 902–918. doi:10.1037/a0025199.
746 Doosje, B., Ellemers, N., & Spears, R. (1995). Perceived intragroup variability as a function of group
747 status and identification. Journal of Experimental Social Psychology, 31, 410–436. doi:10.1006/jesp.
748 1995.1018.
749 Dyer, J., & McWhinnie, S. (2011). A survey of chemistry and physics postdoctoral researchers’
750 experiences and career intentions. International Journal of Gender, Science, and Technology, 3,
751 596–619.
752 Eagly, A. H., Wood, W., & Diekman, A. B. (2000). Social role theory of sex differences and similarities:
753 A current appraisal. In T. Eckes & H. M. Trautner (Eds.), The developmental social psychology of
754 gender (pp. 123–174). Mahwah, NJ: Erlbaum.
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
755 Eccles, J.S. (1989). Bringing young women to math and science. In M. Crawford & M. Gentry (Eds.)
756 Gender and thought: Psychological perspectives (pp. 36–58). New York: Springer. doi:10.1007/
757 978-1-4612-3588-0_3
758 Eccles, J. S. (2005). Subjective task value and the Eccles et al. model of achievement-related choices. In
759 A. J. Elliot & C. S. Dweck (Eds.). Handbook of competence and motivation (pp. 105–121) New
760 York, NY: The Guilford Press.
761 Eccles, J. S. (2009). Who am I and what am I going to do with my life? Personal and collective identities
762 as motivators of action. Educational Psychologist, 44, 78–89. doi:10.1080/00461520902832368.
763 Eccles, J. S., Barber, B., & Jozefowicz, D. (1999). Linking gender to educational, occupational, and
764 recreational choices: Applying the Eccles et al. model of achievement-related choices. In W.
765 B. Swann, Jr., J. H. Langlois, & L. A. Gilbert (Eds.), Sexism and stereotypes in modern society: The
766 gender science of Janet Taylor spence (pp. 153–192). Washington, DC: American Psychological
767 Association. doi:10.1037/10277-007
768 Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual Review of
769 Psychology, 53, 109–132. doi:10.1146/annurev.psych.53.100901.135153.
770 European Commission. (2012). She figures 2012: Gender in research and innovation. Retrieved
771 November 2014 from http://ec.europa.eu/research/science-society/document_library/pdf_06/she-
772 figures-2012_en.pdf.
773 Fiske, S. T., Cuddy, A. J. C., & Glick, P. (2007). Universal dimensions of social cognition: Warmth and
774 competence. TRENDS in Cognitive Sciences, 11, 77–83. doi:10.1016/j.tics.2006.11.005.
775 Forbes, C. E., Schmader, T., & Allen, J. J. B. (2008). The role of devaluing and discounting in
776 performance monitoring: A neurophysiological study of minorities under threat. SCAN, 3, 253–261.
777 doi:10.1093/scan/nsn012.
778 Glynn, S. M., & Koballa, T. R. (2006). Motivation to learn in college science. In J. Mintzes & W.
779 H. Leonard (Eds.), Handbook of college science teaching (pp. 25–32). Arlington, VA: National
780 Science Teachers Association Press.
781 Glynn, S. M., Taasoobshirazi, G., & Brickman, P. (2009). Science Motivation Questionnaire: 
782 Construct validation with nonscience majors. Journal of Research in Science Teaching, 46, 127-146.
783
784
doi: 10.1002/tea.20267.
Good, C., Rattan, A., & Dweck, C. S. (2012). Why do women opt out? Sense of belonging and women’s
785 representation in mathematics. Journal of Personality and Social Psychology, 102, 700–717. doi:10.
786 1037/a0026659.
787 Hammond, R. (2009). Chien-Shiung Wu: Pioneering nuclear physicist (makers of modern science). New
788 York, NY: Chelsea House Publications.
789 Harackiewicz, J. M., Rozek, C. S., Hulleman, C. S., & Hyde, J. S. (2012). Helping parents to motivate
790 adolescents in mathematics and science: An experimental test of a utility-value intervention.
791 Psychological Science, 23, 899–906. doi:10.1177/0956797611435530.
792 Helgeson, V. S. (1994). The relation of agency and communion to well-being: Evidence and potential
793 explanations. Psychological Bulletin, 116, 412–428.
794 Hulleman, C. S., Godes, O., Hendricks, B. L., & Harackiewicz, J. M. (2010). Enhancing interest and
795 performance with a utility value intervention. Journal of Educational Psychology, 102, 880–895.
796 doi:10.1037/a0019506.
797 Hulleman, C. S., & Harackiewicz, J. M. (2009). Promoting interest and performance in high school
798 science classes. Science, 326, 1410–1412. doi:10.1126/science.1177067.
799 International Union of Pure and Applied Physics. (June, 2000). Women in physics working group.
800 Retrieved November 2014 from http://www.if.ufrgs.br/iupap/report1.html.
801 Inzlicht, M., & Ben-Zeev, T. (2000). A threatening intellectual environment: Why females are susceptible
802 to experiencing problem-solving deficits in the presence of males. Psychological Science, 11,
803 365–371. doi:10.1111/1467-9280.00272.
804 Jacobs, J. E., Davis-Kean, P., Bleeker, M., Eccles, J. S., & Malanchuk, O. (2005). ‘‘I can, but I don’t want
805 to’’: The impact of parents, interest, and activities on gender differences in math. In A. M. Gallagher
806 & J. C. Kaufman (Eds.), Gender differences in mathematics: An integrative psychological approach
807 (pp. 246–263). New York, NY: Cambridge University Press.
808 Jacobs, J. E., & Eccles, J. S. (1992). The impact of mothers’ gender-role stereotypic beliefs on mothers’
809 and children’s ability perceptions. Journal of Personality and Social Psychology, 63, 932–944.
810 doi:10.1037/0022-3514.63.6.932.
811 Karpinski, A., & Hilton, J. L. (2001). Attitudes and the implicit association test. Journal of Personality
812 and Social Psychology, 81, 774–788. doi:10.1037/0022-3514.81.5.774.
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
813
814
815
816 Keller, J., & Sekaquaptewa, D. (2008). Solo status and women’s spatial test performance: The role of
817 individuation tendencies. European Journal of Social Psychology, 38, 1044–10553. doi:10.1002/
818 ejsp.490.
819 Konrad, A. M., Ritchie, J. E., Lieb, P., & Corrigall, E. (2000). Sex differences and similarities in job
820 attribute preferences: A meta-analysis. Psychological Bulletin, 126, 593–641. doi:10.1037/0033-
821 2909.126.4.593.
822 Lawrence, J. S., & Crocker, J. (2009). Academic contingencies of self-worth impair positively- and
823 negatively-stereotyped students’ performance in performance-goal settings. Journal of Research in
824 Personality, 43, 868–874. doi:10.1016/j.jrp.2009.05.002.
825 Lent, R. W., Hackett, G., & Brown, S. D. (1996). A social cognitive framework for studying career choice
826 and transition to work. Journal of Vocational Education Research, 21, 3–31.
827 Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self evaluation of one’s social
828 identity. Personality and Social Psychology Bulletin, 18, 302–318. doi:10.1177/0146167292183006.
829 Maddux, W. W., & Brewer, M. B. (2005). Gender differences in relational and collective bases for trust.
830 Group Processes & Intergroup Relations, 8, 159–171. doi:10.1177/1368430205051065.
831 Markus, H. (1977). Self-schemas and processing information about the self. Journal of Personality and
832 Social Psychology, 35, 63–78. doi:10.1037/0022-3514.35.2.63.
833 Martinez, E. D., Botos, J., Dohoney, K. M., Geiman, T. M., Kolla, S. S., Olivera, A., et al. (2007). Falling
834 off the academic bandwagon. Women are more likely to quit at the postdoc to principal investigator
835 transition. EMBO Reports, 8, 977–981. doi:10.1038/sj.embor.7401110.
836 Marx, D. M., & Goff, P. A. (2005). Clearing the air: The effect of experimenter race on target’s test
837 performance and subjective experience. British Journal of Social Psychology, 44, 645–657. doi:10.
838 1348/014466604X17948.
839 McGee, R., Jr, & Keller, J. L. (2007). Identifying future scientists: Predicting persistence into research
840 training. CBE Life Science Education, 6, 316–331. doi:10.1187/cbe.07-04-0020.
841 McGee, R., Jr, Saran, S., & Krulwich, T. A. (2012). Diversity in the biomedical research workforce:
842 Developing talent. Mount Sinai Journal of Medicine, 79, 397–411. doi:10.1002/msj.21310.
843 McKinsey & Company. (2007).Women matter: Gender diversity, a corporate performance driver. http://
844 www.mckinsey.com/features/women_matter.
845 Morgan, C., Isaac, J. D., & Sansone, C. (2001). The role of interest in understanding the career choices of
846 female and male college students. Sex Roles, 44, 295–320. doi:10.1023/A:1010929600004.
847 Murphy, M. C., Steele, C. M., & Gross, J. J. (2007). Signaling threat: How situational cues affect women
848 in math, science, and engineering settings. Psychological Science, 18, 879–885. doi:10.1111/j.1467-
849 9280.2007.01995.x.
850 Muthen, L. K., & Muthen, B. O. (2012). Mplus: The comprehensive modeling program for applied
851 researchers [Computer software]. Los Angeles: Author.
852 National Science Board. (2012). Science and engineering indicators 2012. NSB 10-01. Arlington, VA:
853 National Science Foundation.
854 National Science Foundation, National Center for Science and Engineering Statistics. (2011). Science &
855 engineering degrees: 1966–2008. Detailed statistical tables NSF 11-136. Arlington, VA. http://
856 www.nsf.gov/statistics/nsf11316/pdf/nsf11316.pdf.
857 Nature Neuroscience. (2010). Wanted: Women in research. Nature Neuroscience, 13, 267. doi:10.1038/
858 nn0310-267.
859 Nguyen, H. H. H., & Ryan, A. M. M. (2008). Does stereotype threat affect test performance of minorities
860 and women? A meta-analysis of experimental evidence. Journal of Applied Psychology, 93,
861 1314–1334. doi:10.1037/a0012702.
862 Nosek, B. A., Banaji, M. R., & Greenwald, A. G. (2002). Math=male, me=female, therefore math=me.
863 Journal of Personality and Social Psychology, 83, 44–59. doi:10.1037//0022-3514.83.1.44.
864 Nosek, B. A., et al. (2009). National differences in gender-science stereotypes predict national sex
865 differences in science and math achievement. PNAS, 106, 10593–10597. doi:10.1073/pnas.
866 0809921106.
867 Nye, C. D., Su, R., Rounds, J., & Drasgow, F. (2012). Vocational interests and performance: A
868 quantitative summary over 60 years of research. Perspectives on Psychological Science, 7, 384–403.
869 doi:10.1177/1745691612449021.
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
870 Osborne, J. W., & Jones, B. D. (2011). Identification with academics and motivation to achieve in school:
871 How the structure of the self influences academic outcomes. Educational Psychology Review, 23,
872 131–158. doi:10.1007/s10648-011-9151-1.
873 Osborne, J. W., & Walker, C. (2006). Stereotype threat, identification with academics, and withdrawal
874 from school: Why the most successful students of colour might be most likely to withdraw.
875 Educational Psychology, 26, 563–577. doi:10.1080/01443410500342518.
876 Oswald, D. L., & Harvey, R. D. (2000). Hostile environments, stereotype threat, and math performance
877 among undergraduate women. Current Psychology: Developmental, Learning, Personality, Social,
878 4, 338–356. doi:10.1007/s12144-000-1025-5.
879 Pohlmann, K. (2001). Agency- and communion-orientation in life goals: Impacts on goal pursuit
880 strategies and psychological wellbeing. In P. Schmuck & K. M. Sheldon (Eds.), Life goals and well-
881 being: Towards a positive psychology of human striving (pp. 68–84). Seattle, WA: Hogrefe and
882 Huber.
883 Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple
884 mediation models. Behavior Research Methods, Instruments, & Computers, 36, 717–731. doi:10.
885 3758/BF03206553.
886 President’s Council of Advisors on Science and Technology. (2012, February). Report to the president:
887 Engage to excel: Producing one million additional college graduates with degrees in science,
888 technology, engineering, and mathematics. http://www.whitehouse.gov/sites/default/files/
889 microsites/ostp/pcast-engage-to-excel-final_feb.pdf.
890 Pronin, E., Steele, C. M., & Ross, L. (2003). Identity bifurication in response to stereotype threat: Women
891 and mathematics. Journal of Experimental Social Psychology, 40, 152–168. doi:10.1016/S0022-
892 1031(03)00088-X.
893 Rahm, J., & Charbonneau, P. (1997). Probing stereotypes through students’ drawings of scientists.
894 American Journal of Physics, 65, 774–778. doi:10.1119/1.18647.
895 Ramsey, L. R., Betz, D. E., & Sekaquaptewa, D. (2013). The effects of an academic environment
896 intervention on science identification among women in STEM. Social Psychology of Education, 16,
897 377–397. doi:10.1007/s11218-013-9218-6.
898 Schmader, T. (2002). Gender identification moderates stereotype threat effects on women’s math
899 performance. Journal of Experimental Social Psychology, 38, 194–201. doi:10.1006/jesp.2001.
900 1500.
901 Schmader, T., Johns, M., & Barquissau, M. (2004). The costs of accepting gender differences: The role of
902 stereotype endorsement in women’s experience in the math domain. Sex Roles, 50, 835–850. doi:10.
903 1023/B:SERS.0000029101.74557.a0.
904 Sekaquaptewa, D., & Thompson, M. (2003). Solo status, stereotype threat, and performance expectancies:
905 Their effects on women’s performance. Journal of Experimental Social Psychology, 39, 68–74.
906 doi:10.1016/S0022-1031(02)00508-5.
907 Seymour, E., & Hewitt, N. M. (1995). Talking about leaving: Why undergraduates leave the sciences.
908 Boulder, CO: Westview Press.
909 Shapiro, J. R., & Williams, A. M. (2012). The role of stereotype threats in undermining girls’ and
910 women’s performance and interest in STEM fields. Sex Roles, 66, 175–183. doi:10.1007/s11199-
911 011-0051-0.
912 Smith, J. L. (2004). Understanding the process of stereotype threat: A review of mediational variables and
913 new performance goal directions. Educational Psychology Review, 16, 177–206. doi:10.1023/B:
914 EDPR.0000034020.20317.89.
915 Smith, J. L., Cech, E., Metz, A., Huntoon, M., & Moyer, C. (2014a). Giving back or giving up: Native
916 American student experiences in science and engineering. Cultural Diversity and Ethnic Minority
917 Psychology, 20, 413–429. doi:10.1037/a0036945.
918 Smith, C. S., & Hung, L. C. (2008). Stereotype threat: Effects on education. Social Psychology of
919 Education, 11, 243–257.
920 Smith, J. L., Lewis, K. L., Hawthorne, L., & Hodges, S. D. (2013). When trying hard isn’t natural:
921 Women’s belonging with and motivation for male-dominated STEM fields as a function of effort
922 expenditure concerns. Personality and Social Psychology Bulletin, 39, 131–143. doi:10.1177/
923 0146167212468332.
924 Smith, J. L., Thoman, D. B., Deemer, E. D., & Zazworsky, L. (2014b). Motivation under the microscope:
925 Understanding undergraduate science students’ multiple motivations for research. Motivation and
926 Emotion, 38, 496–512. doi:10.1007/s11031-013-9388-8.
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
927 Smith, J. L., & White, P. H. (2001). Development of the domain identification measure: A tool for
928 investigating stereotype threat effects. Educational and Psychological Measurement, 61,
929 1040–1057. doi:10.1177/00131640121971635.
930 Smith, J. L., & White, P. H. (2002). An examination of implicitly activated, explicitly activated, and
931 nullified stereotypes on mathematical performance: It’s not just a woman’s issue. Sex Roles, 47,
932 179–191. doi:10.1023/A:1021051223441.
933 Spencer, S. J., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and women’s math performance.
934 Journal of Experimental Social Psychology, 35, 4–28. doi:10.1006/jesp.1998.1373.
935 Stangor, C., Carr, C., & Kiang, L. (1998). Activating stereotypes undermines task performance
936 expectations. Journal of Personality and Social Psychology, 75, 1191–1197. doi:10.1037/0022-
937 3514.75.5.1191.
938 Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance.
939 American Psychologist, 52, 613–629. doi:10.1037/0003-066X.52.6.613.
940 Steele, J., James, J. B., & Barnett, R. (2002a). Learning in a man’s world: Examining the perceptions of
941 undergraduate women in male-dominated academic areas. Psychology of Women Quarterly, 26,
942 46–50. doi:10.1111/1471-6402.00042.
943 Steele, C. M., Spencer, S. J., & Aronson, J. (2002b). Contending with group image: The psychology of
944 stereotype and social identity threat. Advances in Experimental Social Psychology, 34, 379–440.
945 doi:10.1016/S0065-2601(02)80009-0.
946 STEMconnector. (2012). Where are the STEM students? What are their career interests? Where are the
947 STEM jobs? http://www.stemconnector.org/sites/default/files/store/STEM-Students-STEM-Jobs-
948 Executive-Summary.pdf.
949 Stout, J. G., Dasgupta, N., Hunsinger, M., & McManus, M. A. (2011). STEMing the Tide: Using ingroup
950 experts to inoculate women’s self-concept in science, technology, engineering, and mathematics
951 (STEM). Journal of Personality and Social Psychology, 100, 255–272. doi:10.1037/a0021385.
952 Su, R., Rounds, J., & Armstrong, P. I. (2009). Men and things, women and people: A meta-analysis of sex
953 differences in interests. Psychological Bulletin, 135, 859–884. doi:10.1037/a0017364.
954 Swann, W. B., Jr, & Bosson, J. K. (2010). Self and Identity. In S. T. Fiske, D. T. Gilbert, & G. Lindzey
955 (Eds.), Handbook of social psychology (5th ed., pp. 589–628). New York: McGraw-Hill.
956 Thoman, D. B., Smith, J. L., Brown, E. R., Chase, J., & Lee, J. Y. K. (2013). Beyond performance: A
957 motivational experiences model of stereotype threat. Educational Psychology Review, 25, 211–243.
958 doi:10.1007/s10648-013-9219-1.
959 Thoman, D. B., White, P. H., Yamawaki, N., & Koishi, H. (2008). Variations of gender-math stereotype
960 content affect women’s vulnerability to stereotype threat. Sex Roles, 58, 702–712. doi:10.1007/
961 s11199-008-9390-x.
962 Tine, M., & Gotlieb, R. (2013). Gender-, race-, and income-based stereotype threat: The effects of
963 multiple stigmatized aspects of identity on math performance and working memory function. Social
964 Psychology of Education, 16, 353–376. doi:10.1007/s11218-013-9224-8.
965 Urry, M. (July, 2002). The status of women in physics—an international meeting on what, why, and how
966 to change. American Physical Society. Retrieved November 2014 from http://www.aps.org/units/
967 fps/newsletters/2002/july/urry.pdf.
968 Walsh, P. E., & Smith, J. L. (2007). Opposing standards within the cultural worldview: Terror
969 management and American women’s desire for uniqueness versus inclusiveness. Psychology of
970 Women Quarterly, 31, 103–113. doi:10.1111/j.1471-6402.2007.00335.x.
971 Wang, M., & Degol, J. (2013). Motivational pathways to STEM career choices: using expectancy-value
972 perspective to understand individual and gender differences in STEM fields. Developmental Review,
973 33, 304–340. doi:10.1016/j.dr.2013.08.001.
974 Webb, R. M., Lubinski, D., & Benbow, C. P. (2002). Mathematical facile adolescents with math-science
975 aspirations: New perspectives on their educational and vocational development. Journal of
976 Educational Psychology, 94, 785–794. doi:10.1037/0022-0663.94.4.785.
977 Weisgram, E. S., Bigler, R. S., & Liben, L. S. (2010). Gender, values, and occupational interests among
978 children, adolescents, and adults. Child Development, 81, 778–796. doi:10.1111/j.1467-8624.2010.
979 01433.x.
980
981
Wickware, P. (1997). Along the leaky pipeline. Nature, 390, 202–203. doi:10.1038/36639.
982
Women’s science identity
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f
U
N
C
O
R
R
E
C
T
E
D
P
R
O
O
F
983 Woodcock, A., Hernandez, P. R., Estrada, M., & Schultz, P. W. (2012). The consequences of chronic
984 stereotype threat: Domain disidentification and abandonment. Journal of Personality and Social
985 Psychology, 103, 635–646. doi:10.1037/a0029120.
986 Xie, Y., & Shauman, K. A. (2003). Women in science: Career processes and outcomes. Cambridge, MA:
987 Harvard University Press.
988 Young, D. M., Rudman, L. A., Buettner, H. M., & McLean, M. C. (2013). The influence of female role
989 models on women’s implicit science cognitions. Psychology of Women Quarterly, 37, 283–292.
990 doi:10.1177/036168431348210.
Jessi L. Smith received her Ph.D. from the University of Utah in 2002 and is now a Professor of
Psychology and Special Assistant to the Provost at Montana State University in Bozeman Montana, USA.
Her research examines the social psychology of motivation, particularly as a function of cultural norms
about gender, race and ethnicity, and sexuality. Her work has been applied to educational, career,
political, and health related domains.
Elizabeth R. Brown is now an Assistant Professor in the Department of Psychology at the University of
North Florida in Jacksonville Florida, USA. Her research focuses on motivation and support for the
current sociopolitical system, with an emphasis on how gender, culture, and stereotypes shape these
processes.
Dustin B. Thoman is an Associate Professor in the Department of Psychology at California State
University Long Beach. His primary research focus is on understanding how and why people develop
interest and sustain motivation for specific academic domains, careers, and other lifelong pursuits. He
tests how these motivational experiences and self-regulation processes can be supported or disrupted by
the social and cultural context in which interests are sparked, developed, and ultimately integrated (or not)
with one’s identity and self-concept.
Eric D. Deemer is an Assistant Professor in the Department of Educational Studies at Purdue University
in West Lafayette, Indiana USA. He received his Ph.D. in counseling psychology from the University at
Albany, State University of New York in 2008. His research focuses broadly on STEM career
development, with related interests in academic achievement motivation, stereotype threat, and research
training in professional psychology.
J. L. Smith et al.
123
Journal : Small-ext 11218 Dispatch : 3-2-2015 Pages : 24
Article No. : 9296 * LE * TYPESET
MS Code : SPOE-D-14-00082 R CP R DISK
A
u
th
o
r
 P
r
o
o
f