Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Sep 1.
Published in final edited form as: J Res Adolesc. 2016 Oct 12;27(3):492–505. doi: 10.1111/jora.12289

Shifting STEM Stereotypes? Considering the Role of Peer and Teacher Gender

Catherine Riegle-Crumb 1, Chelsea Moore 2, Jenny Buontempo 3
PMCID: PMC5546141  NIHMSID: NIHMS816701  PMID: 28776845

Introduction

Despite the educational progress of women in the last fifty years and the fact that they currently outnumber men in rates of college matriculation and graduation nationwide, women remain under-represented in many science, technology, engineering, and math (STEM) fields (Diprete & Buchmann 2013; Gerber & Cheung 2008). Such continued disparities are particularly troubling because they lead to subsequent inequality, including occupational segregation and gender gaps in income and status (Shauman, 2006). While there has been some movement towards gender parity in past decades, particularly in the biological sciences, desegregation appears to be currently stalled (England, 2010). The strongest patterns of inequality are found in the field of engineering where women comprise less than twenty percent of bachelor degree earners (U.S. Department of Education, 2012).

For decades, scholars have called attention to the role of stereotypes in producing and maintaining segregation, such that exposure to messages about females’ supposed natural inferiority in math and related domains can negatively influence their attitudes in such fields, including dampening their confidence in their ability and their interest in pursuing fields where others assume they are not skilled or do not belong (Charles & Bradley 2002; Correll, 2001; Eccles, 1994; Oakes, 1990; Robnett & Leaper, 2013; Seymour & Hewitt, 1997). While mostly experimental in nature, a host of studies on stereotype threat have provided compelling new empirical evidence regarding the impact of stereotypes on females’ attitudes and subsequent STEM-related choices (Lesko & Corpus, 2006; Oswald, 2008; Smith, 2006). Correspondingly, recent research has explored the effectiveness of specific factors or interventions aimed at preventing the negative consequences of exposure to stereotypes; this includes priming alternative identities (e.g. academic identity as a good student), giving women new resources to resist stereotypes (such as teaching a growth mindset), or providing tangible evidence that contradicts stereotypes (Dasgupta & Asgari, 2004; Gresky et al., 2005). Such research offers important insights into how females can thrive in the face of obstacles, and yet at the same time it only addresses one part of the larger problem of inequality.

Specifically, segregation is maintained not only by the actions of women (deciding not to enter a STEM field or not stay in it), but also by the actions of men. Gender stereotypes clearly favor men and embolden their choices, giving them greater confidence in their abilities and their ‘natural’ inclination towards such fields (Ridgeway & Correll, 2004). By definition men greatly outnumber women in male-dominated STEM fields, and therefore the ‘chilly climate’ or negative atmosphere of male-dominated classrooms as documented at the college level is largely created and maintained by men (Morris & Daniel, 2008). Biased views of women’s abilities among the mostly male STEM faculty in both the academy and industry have been linked to discrimination against women in the hiring process, as well as on the job (Bobbitt-Zeher, 2011; Ecklund, Lincoln & Tansey, 2012). And yet despite these incontrovertible realities, current research pays little attention to male students’ views and beliefs about females, and more importantly, whether and how negative views can be changed.

To address this critical shortage in the literature, we utilize high school engineering classes as a strategic research site to investigate the views of contemporary adolescent males who have already indicated a strong interest in pursuing STEM fields, and thus who will likely comprise future STEM college classrooms and eventually occupations. Our study builds on gender theories about the power of counter-stereotypical evidence in the form of female peers and adult role models for improving young females’ STEM attitudes and outcomes (Dasgupta & Asari, 2004), but instead considers how such evidence might impact males’ stereotypical views. Specifically, informed by the lens of intergroup contact theory (Pettigrew & Tropp, 2006), we suggest that sustained exposure to a female teacher and female peers in a high school engineering classroom over the course of a year could lead young males to decrease their gender stereotypical views about STEM ability. Importantly, we advance prior literature on gender inequality as well as intergroup contact by considering whether the initial strength of males’ stereotypical views has implications for how responsive they are to such gender cues, as those who enter the classroom with the most biased viewpoints are likely to react differently to the presence of female peers and/or a female teacher compared to those with less biased viewpoints (Flynn, 2005).

Our study is comprised of several hundred high school students in a new high school engineering course. Our analyses model the change in boys’ views over the course of the year in relation to the gender of their teacher and the percentage of female peers in the classroom, net of their individual and school characteristics. Finally, we note that while we also empirically examine females’ views, the primary focus and contribution of this study is on patterns for young men. Rather than suggest that young women need to ‘lean in’ to the norms and culture of a male-dominated environment, we are instead interested in unpacking whether and how the stereotypical views of young males can change. In doing so, we hope to contribute new knowledge that can ultimately change the experiences of future generations of females by reducing the chances that they will have to encounter gender/STEM stereotypes throughout their educational and occupational trajectories.

Background

Who believes the hype?

The empirical literature on gender/STEM stereotypes finds that, not surprisingly, females have relatively low levels of endorsement of explicit statements about men’s presumed natural superiority or inclination towards these fields (Blanton, Christie & Dye, 2002; Bonnot & Croizet, 2007; Cvencek, Meltzoff, & Greenwald, 2011; Young et al., 2013), although studies measuring implicit or unconscious stereotypes find somewhat stronger evidence (Lane, Goh, & Driver-Linn, 2012; Nosek, Banaji, & Greenwald, 2002). In particular, several recent studies have found that young women who have chosen STEM fields in college are very unlikely to endorse such stereotypes (Kiefer & Sekaquaptewa, 2007; Schmader, Jones & Narquissau, 2004; Smeding, 2012), which is logical as their choices and achievement largely contradict such views. Of course, this is not to suggest that females in STEM fields are necessarily immune to the negative effects of stereotypical viewpoints of others; indeed research has shown that their confidence, interest, and performance often suffers from such exposure, ultimately lowering the probability that they continue to persist (Correll, 2001; Smith, 2006).

The literature on males’ endorsement of gender/STEM stereotypes is comparatively limited, but offers evidence that men are generally more likely to endorse such stereotypes than women, and that in particular, men in STEM fields are very likely to hold gender stereotypical views (Nosek et al., 2002; Smeding, 2012; Young et al., 2013). Again, these patterns make sense, particularly as men who are pursuing STEM benefit from the stereotype that they are naturally gifted and prone to be successful in their chosen field; indeed this belief can help buffer negative psychological effects of any instance of low performance, as poor performance can be attributed to bad luck or extenuating circumstances, and not their actual ability (Bench et al., 2015; Niederle & Vesterlund, 2010). Yet given that males are arguably the most likely source of females’ exposure to stereotypes, it is critical to expand upon the limited research in this area to better understand males’ beliefs.

This study will address this limitation and provide new information not only regarding the views of a contemporary cohort of young males who are inclined to pursue STEM in their adult futures, but also regarding the extent to which they might change over time. While gender scholars have noted that stereotypes are no doubt resilient and persistent (Reskin, 2000; Ridgeway, 1997), we outline an argument for why we might expect change under certain conditions. First, we briefly discuss research that highlights how gendered cues provided by peers and adult role models can shape girls’ STEM related outcomes. Second, we connect the insights of this female-focused research to the theory of intergroup contact, and subsequently argue that similar gender cues could in fact impact boys, ultimately decreasing their stereotypical views.

The benefits of female peers and adult role models for girls in STEM

Sociological theories of gender as a social construct argue that while gender is a multi-level system, there is also the potential for micro-level environments to subvert larger paradigms or norms (Ridgeway, 1997; Ridgeway & Correll, 2004). Empirical evidence supports this, finding that gendered aspects of local environments matter. Specifically, exposure to female peers and adult role models who themselves defy gender/STEM stereotypes has been shown to be a positive influence on girls’ STEM-related attitudes and choices in high school and through the early years of college (Dasgupta & Asgari, 2004; Lenton, Bruder, & Sedikides, 2009; Stout et al, 2011). And while most females are unlikely to endorse STEM stereotypes themselves (as discussed earlier), the counter-stereotypical evidence provided by accomplished female peers and adults in STEM fields can play a particularly powerful role as girls move through adolescence and grapple with the increasing saliency of future role expectations (Eccles, 2015; Evans, Whigham, & Wang, 1995; Stout et al., 2011). Furthermore, while brief exposure to such same-sex individuals can be beneficial, it is logically longer durations and more engaged interaction that most effectively bolsters girls’ STEM-related confidence, ambitions, and related choices (Blanton, Crocker, & Miller 2000; Stout et al., 2011; Young et al., 2013). In sum, this body of research has provided important theory and corresponding evidence regarding the benefits of same-gender peers and adult role models in shaping girls’ views from early through late adolescence; yet at the same time it has virtually ignored the potential for impact on boys. In the sections below, we address this oversight in the literature, and argue that female peers and adult experts can be a powerful source of counter-stereotypic evidence for adolescent males.

Why gender composition of STEM environments impact males

Gender scholars have pointed towards the persistence of gender stereotypes in contemporary times, noting that because males benefit from stereotypes, they have a strong incentive to discount any evidence that might undermine such beliefs, treating it as an exception and therefore keeping their belief structure intact (Reskin, 2000; Ridgeway, 1997). Yet in this paper, we suggest that intergroup contact theory offers a useful lens to consider why the stereotypical beliefs of young males regarding females’ ability in STEM fields might change based on gendered aspects of their local environment. This theory is consistent with gender theories about the importance of interaction at the micro-level for either supporting or subverting inequality (Ridgeway, 1997), yet more clearly articulates the factors that facilitate a change in biased beliefs. Originally theorized by Allport (1954) and further developed by other psychologists (Hewstone & Swart, 2011; Pettigrew, 1998), intergroup contact theory predicts that people belonging to the in-group (or privileged group) will become less prejudiced towards an out-group (or less privileged group) under the following optimal conditions of interaction: common status or roles occupied by in-group and out-group members within the situation, common goals achieved through intergroup cooperation, and authority or institutional support. Additionally, Pettigrew (1998) articulates that intergroup contact is most likely to decrease bias and discrimination through extended contact, rather than one-time or brief exposure.

Intergroup contact theory was initially developed to examine how racial/ethnic discrimination could be decreased via interpersonal encounters, and as captured by a recent meta-analysis, there is strong empirical evidence of its positive impact (Pettigrew & Tropp, 2006). Recently scholars have extended their focus to examine the role of intergroup contact in decreasing discriminatory views towards sexual minority individuals (Smith, Axelton, & Saucier, 2009). In contrast, we know of no studies that have applied this theory to how males’ stereotypical or discriminatory viewpoints might change as a consequence of their interaction with females. And while gender stands apart from other axes of stratification between groups (such as race/ethnicity) due to the fact that males and females interact and intermingle continuously in their daily lives, females nonetheless occupy a general position of lower social status compared to males, and this is particularly pronounced within fields that are historically (and currently) dominated by men (Ridgeway, 1997). Thus, we argue in STEM domains that are strongly associated with males, such as engineering classrooms, females do indeed constitute a salient out-group. If repeated intergroup contact occurs in such a context, then there is reason to expect that males’ stereotypical views about female’s STEM-related abilities will be challenged.

Additionally, while intergroup contact theory focuses on the shared status among members, suggesting the importance of peer interaction, it also specifies the importance of the support of authority for such interactions. Authority can be considered at different levels, but within a classroom the teacher is the local authority as well as the expert. If the teacher is also a member of the out-group, in this case female, then it stands to reason that repeated contact in a STEM classroom might also shift boys’ stereotypes. Thus, our study bridges gender theory and intergroup contact theory by positing that while it is in males’ self-interest to believe gender/STEM stereotypes, repeated interactions with female peers or teachers in an environment where such stereotypes are salient or relevant may prove a catalyst for their views to begin to change, particularly during adolescence when youth are sensitive to gendered messages from peers as well as adults within their school context (Eccles & Roeser, 2011; Ridgeway, 1997).

Considering differences in initial beliefs

Finally, this study will provide a new contribution to the literature on both gender inequality and intergroup contact theory by considering how males’ initial level of stereotype endorsement may be associated with whether their views change. Scholars studying the connection between intergroup contact and stereotypes acknowledge that those individuals who are the most prejudiced in their views may resist interaction and not fully engage even in a structured setting, and may also be the most likely to ignore or discount evidence that disconfirms stereotypes, therefore resulting in little change (Flynn, 2005; Pettigrew, 2008). Nonetheless, there is little extant research that empirically considers individual differences in prior viewpoints (Asbrock, Gutenbrunner, & Wagner, 2013; Hewstone & Swart, 2011). This is potentially problematic, as focusing only on change (or the lack thereof) across the entire in-group could easily obscure patterns specific to those with certain viewpoints at the start (Flynn, 2005).

Regarding the particular focus of this study on stereotypic beliefs about gender and STEM, it stands to reason that while males on average may endorse such stereotypes, surely not all of them do. Put briefly, as a consequence of a myriad of differences in prior socialization and experiences, some young men are likely to be more gender-biased than others (Mullen, 2014). We therefore examine the variability in young men’s STEM stereotype endorsements that exists prior to their interaction with peers and teachers who potentially defy such stereotypes, and investigate whether this has implications for subsequent change. In the section that follows, we describe how and why high school engineering classrooms present a strategic research site to investigate these issues.

Current Study

Our study focuses on high school engineering classes, which represent a relatively new advanced course offering available in approximately only ten percent of schools nationwide (Snyder and Dillow, 2011). The typical advanced math and science curriculum for students includes courses such as Chemistry and Pre-calculus. Girls take these classes at equal rates as boys (U.S. Department of Education, 2012); thus the gender composition of these classrooms varies little, and the presence of female students does not constitute counter-stereotypical evidence. Engineering classes, on the other hand, offer an ideal location to study the stereotypical views of young people and how they change; as elective courses that are typically comprised of students interested in pursuing a STEM major in college, high school engineering courses have relatively fewer female students on average (Tai, 2012). Additionally, among STEM fields, engineering remains the educational and occupational field most strongly dominated by men (U.S. Department of Education, 2012). For these reasons, and as alluded to earlier, we suggest that gender is likely to be very salient in high school engineering classrooms, and a greater presence of female students would be considered unusual or potentially disconfirming, as might a female teacher with expertise in this heavily male-dominated field.

Critically for this study, engineering classes are also likely to promote the conditions discussed by scholars of intergroup contact theory as optimal for prompting a change in views. Most high school STEM classrooms could arguably meet the criteria of males and females students sharing a common status as learners in the same class, and include institutional or authority support as evidenced by the fact that it is an official school offering with a teacher. Yet most classes could not be characterized as having shared (rather than individual) goals and a cooperative learning environment. Typical math and science high school classrooms primarily include lectures by teachers and seatwork performed individually by students, thereby not facilitating intergroup contact (Schmidt, 2007). In contrast, high school engineering classes tend to rely on extensive use of group projects, and promote a cooperative learning environment where students work towards solving a common challenge (Tai, 2012). Engineering classes also typically employ inquiry-based learning, where teachers work alongside students to help guide their learning (Marshall & Berland, 2012).

Our specific research site is a high school engineering course offered in nineteen schools across the country. This year-long course was created by engineering and education faculty at a large research university in the Southwest. The summer before they teach the courses, all teachers are trained for a period of several weeks by faculty at the university where the course was developed; further training and support occurs virtually throughout the school year. The summer training includes a session on equity in the classroom, including a focus on gender equality.

The course utilizes a series of design challenges that each span several weeks and require that students learn and apply engineering concepts and design to everyday useful products; examples include a hair dryer and a camera. Collaborative learning is an essential part of the course and students work in teams almost every day to solve these challenges. Teachers utilize a range of different criteria to assign students to different groups throughout the year, including random assignment, alphabetical order of first or last names, or grouping according to favorite musical bands or types of foods. Contact between males and females is further facilitated by the small size of these classes, typically about twenty students.

For the reasons outlined above, this high school engineering course provides an ideal context to examine potential change in males’ endorsement of gender/STEM stereotypes. In classrooms where there are more female students, the opportunities for intergroup contact will be greater than in classrooms with comparatively fewer female students, thereby increasing the likelihood that male students will confront and adjust their stereotypical beliefs over the duration of this year-long course. Additionally, the presence of a female teacher will also increase the occurrence of intergroup contact, potentially making it more likely that male students will hold less stereotypical views. Importantly, as discussed earlier, we anticipate that the impact of female peers and female teachers might differ based on the intensity of students’ initial level of stereotypical views, so we empirically examine this possibility.

Finally, as mentioned above, although this study builds upon and is informed by prior research focusing on the impact of same-sex role models for girls’ STEM attitudes and ambitions, we are interested in using data from current classrooms to focus specifically on boys’ stereotypical views. While we will conduct parallel analyses for girls in our sample, we anticipate that, consistent with literature cited earlier, this group of selective girls in engineering classes is unlikely to endorse gender stereotypes, and therefore anticipate that there is little room to move over the course of the year. And perhaps more importantly, our focus on boys is driven by an interest in highlighting how we can change boys’ views, and in doing so ultimately change the tenor of future STEM college classrooms and occupational sites.

Method

Participants

Data was collected during the 2012–2013 school year. While 23 schools across the country offered the engineering course that year, 19 of them agreed to participate in the research study. Of the participating schools, 52% (n=10) were from the Southwest, 26% (n=5) from the Northeast, 10% (n=2) from the Mid-west, and 10% (n=2) from the West. On average, schools had a student body where 39% qualified for free/reduced lunch (minimum=10%, maximum= 91%). Additionally, the average percent minority (Hispanic or African American) was 46% (ranging from 7% to 99%). Two schools were private, and four schools characterized themselves as STEM-focused. Each of the schools had one teacher assigned to the course. On average, teachers had 10.2 years of teaching experience (minimum=1 and maximum=23). A few teachers (n=5) had engineering undergraduate degrees. Approximately 80% (n=15) of teachers had teaching certification in math or science. Slightly more than one-third of teachers (n=7) were female. We return to the issue of teacher gender below.

Our analytic sample includes 357 students, including 262 males (73%) and 95 females (27%). However, the percent of female students varied across classrooms, as discussed below. Regarding race/ethnicity, white non-Hispanic students comprised the largest group (55%), while 23% of students were Hispanic, 15% were Asian, and 7% were from other racial/ethnic groups (due to small numbers, African American, American Indian and Alaskan Natives students were included in one category). As this course counts as an advanced science credit, it is taken primarily by upperclassmen. Specifically, 78 percent of students enrolled were classified as either juniors or seniors. Table 1 includes additional information about students (overall and separately by gender, discussed below).

Table 1.

Student Characteristics

All Students Male students Female students
Mean or proportion SD Mean or proportion SD Mean or proportion SD
Race/Ethnicity
Non-Hispanic White 0.55 0.58 0.50
Hispanic 0.23 0.23 0.25
Asian 0.15 0.12 0.21*
Other Race/Ethnicity 0.07 0.07 0.04
Parent Education Level 2.85 (1.07) 2.81 (1.06) 2.94 (1.01)
Math Grades
Mostly A’s 0.44 0.46 0.41
Mostly B’s or below 0.56 0.54 0.59
College Expectations
Expects to attend college 0.90 0.88 0.95
Does not expect to attend 0.10 0.12 0.05
Expected College Major
STEM major 0.84 0.88 0.75*
non-STEM major 0.16 0.12 0.25
N 357 262 95
Open in a new tab
*

’s indicate a statistically significant difference between males and females at p<.05

Procedure

Depending on the teachers’ preference, student surveys were administered online (n=13) or on paper (n=6). In the case of paper surveys, students entered their answers onto scantron sheets which were then mailed to the researchers in a previously provided self-addressed envelope. Teachers provided the survey links and/or passed out the paper surveys, but could not see student responses. Students were assured of the confidentiality of their responses on the front page of the survey, which discussed the intent of the survey as gathering information on their experiences and viewpoints. Consent was collected via a signed permission form from both parents and children, and the research team had IRB approval for all activities. Pre-surveys were administered the first weeks of the Fall 2012 school year in order to capture students’ views before taking the course, while post-surveys were given in the last weeks of Spring 2013 near the completion of the course. Our final analytic sample includes 357 students who consented to participate (a response rate of 84%) and who filled out both surveys.

Student Background Measures

Table 1 provides additional information about students’ background as reported in the surveys. Students were asked to report their parents’ education level, which is included here coded as the highest level of education received by either parent, with the following categories: 1= high school degree or less, 2= 2 year or associate degree, 3=college degree, and 4= advanced degree. The average for this measure was 2.85, which falls just short of a college degree. Students were also asked to indicate their average math grades from 9th grade onward, selecting between the categories of mostly A’s (44% of students), mostly B’s (45%), mostly C’s (10%), and mostly D’s or below (less than 1%). Because of the small percentages of students in the last two categories, we created a new dichotomous variable that distinguished between those who made mostly A’s versus those who made mostly B’s or below.

College expectations were measured dichotomously as either expecting to go to a four-year college or university or not, and expectations of majoring in STEM fields were similarly coded. Almost all students reported an expectation of going to college (90%), and intentions to major in STEM were also very high, although significantly more males (88 percent) than females (75 percent) reported such expectations.

Measures of the Gendered Environment of the Classroom

Our two primary independent variables of interest include teachers’ and students’ gender. To capture the gender composition of students in the classroom, we created a continuous variable that measured the percentage of females enrolled in the engineering course, with an overall classroom mean of 24 percent and a range from a low of 6 percent to a high of 59 percent. Girls were somewhat clustered within classrooms, as the average female student was in a classroom that was 34 percent female and the average male was in a classroom that was 20 percent female (t = 7.90, p-value = .000). For teacher gender, we created a dichotomous variable coded 1 for female teachers and 0 for male. On average, almost twenty percent of students had a female teacher; however we note that female students (30 percent) were significantly more likely than male students (16 percent) to have a female teacher (t = 2.94, p-value = .003). Consequently there is a modest correlation between percent female in the classroom and the presence of a female teacher (.34), suggesting that perhaps female students were more inclined to enroll in this elective course when the teacher was also female.

Measures of Gender Stereotypes

Our survey included several questions that measure students’ explicit endorsement of gender/STEM stereotypes, taken from the Michigan Study of Adolescent and Adult Life Transitions (MSALT), a large longitudinal study that includes a range of questions about gender attitudes and beliefs (Bleeker & Jacobs, 2004; Eccles, 2015; Eccles & Harold, 1991). Specifically, we included items to capture views about gender differences in both mathematical and mechanical aspects that are critical components of engineering (National Academy of Engineering, 2008; Pawley 2009). This included the following three items: 1) men are naturally better at advanced math than women; 2) men are naturally better at mechanical things than women; and 3) men find math more useful than women. Categories of response were: 1 (strongly disagree), 2 (disagree), 3 (neutral), 4(agree), and 5 (strongly agree). We created a scale that averages students’ responses to these questions, noting that a high value on the scale indicates a more gender stereotypical view. The alpha reliability of the scale was very high, approximately .9.

Not surprisingly, female engineering students reported significantly lower levels of endorsement of this gender/STEM stereotype than male engineering students at the beginning of the year (t=7.10, p-value=.000). Girls’ mean at time 1 was 2.04, indicating that on average girls disagreed with gender/STEM stereotypes. Boys’ mean at time 1 was 2.81, indicating that on average they were neutral. A similar disparity existed at the end of the year (girls’ mean=1.97, boys’ mean=2.83; t=8.15, p-value=.000). To measure individuals’ change (or lack thereof) over the course of the year, we created a difference score by subtracting students’ time 1 score from their time 2 score. This serves as the dependent variable in our analyses, such that a positive score indicates an increase in stereotypical views and a negative score indicates a decrease in such views.

Analytic Strategies

The goal of our analyses is to examine whether the presence of more female peers or a female teacher in high school engineering classes predicts a change in male (or female) students’ stereotypical views over the course of a year. Because we are interested in examining whether such changes depend on students’ initial level of stereotype endorsement, we fit models with main effects for percent female and teacher gender (neither were significant), and then added interaction terms between students’ time 1 stereotype score and the measures for percent female and teacher gender, respectively. We first fit models with interaction terms using the original linear version of the time 1 scale, and then also fit models with a categorical version of students’ time 1 score that allowed us to test for non-linear relationships. Specifically, we created a categorical version that made a substantive distinction between those who initially rejected the stereotype (defined as an average score <2.5), those who were neutral about it (average score >=2.5 and <3.5), and those who endorsed the stereotype (score >=3.5) at the beginning of the year (we discuss the robustness of our results to different category thresholds later). Results revealed that while the interactions between both percent female and teacher gender with the linear time 1 score were not significant, interaction terms with the categorical version were indeed statistically significant (and model fit indicators also preferred the categorical version).

For ease of interpretation we choose to show results separately for the three categories of students (rejecter, neutral, and endorser) to clearly reveal the impact (or lack thereof) of female peers and/or a female teacher on the change in students’ stereotypical views according to their initial level of endorsement. We also ran analyses separately by gender (and accordingly discuss the results of tests of differences in coefficients across models between groups and between genders when appropriate). We note that male students were relatively equally divided among the 3 categories (36% in the rejecter category, 35% in the neutral category, and 28% in the endorser category), while female students fell into only the rejecter category (72%) and the neutral category (28%).

To ensure that any associations we observe are robust to differences in students’ academic and social background, models include the student characteristics shown in Table 1; however given the very limited variation in college expectations and intentions of majoring in STEM (and the fact that these measures were not statistically significant and did not alter key findings) we chose not to include those measures in the final models presented here. We also included measures for the percent minority in the school as well as a measure for private schools (exploratory analyses using additional school as well as teacher measures produced null results and are not included here). Missing data was found only for parent education level (41 students), which was then imputed via single imputation. Because students are nested within classrooms, all results shown are from hierarchical linear random effects models (because only school offered more than one section of the engineering classroom, our results did not change whether we specified classrooms or schools as the level two unit). Unconditional models confirmed that there was statistically significant variation for males from all three groups, with intraclass correlations (ICC) between .04–.06. However, the intraclass correlations for both groups of girls were only .01 and not statistically significant; for the sake of parallel models we choose to use multi-level models for both genders. All analyses were performed using the statistical software, Stata.

Results

Results for Male students

Table 2 displays the results of our multivariate models for male students. As seen in the first column, among male students who initially reject the stereotype of male superiority, the percentage of females in the classroom has a significant and negative effect. Recalling that a low (negative) score on the dependent variable indicates a decrease in stereotypical views about males’ STEM superiority, this indicates that among this group of male students who generally disagree with the stereotype, having more females in the classroom results in them having even lower stereotype scores at the end of the year. We find no significant effect regarding the gender of the teacher. Finally, regarding other student and school characteristics, male students with higher levels of parent education and those who attend schools with a higher percentage of minority students also significantly decrease their stereotypical views over the course of the year, while high math grades and private school attendance significantly predict an increase in stereotypical views.

Table 2.

Predicting Change in Gender/STEM Stereotypes for Male Students by their Initial Level of Endorsement

Rejecters at Time 1 Neutral at Time 1 Endorsers at Time 1
Gendered Environment of Classroom
Female peers (percent) −2.011** (0.772) −0.706 (0.731) −0.035 (0.889)
Female teacher 0.435 (0.301) 0.212 (0.366) −0.885** (0.348)
Race/Ethnicity (reference: non-Hispanic white)
Hispanic −0.162 (0.241) 0.126 (0.263) 0.677* (0.281)
Asian 0.038 (0.296) 0.273 (0.252) −0.385 (0.377)
Other race/ethnicity −0.229 (0.318) 0.184 (0.311) −0.211 (0.565)
Parent education level −0.218* (0.087) 0.029 (0.088) 0.295* (0.129)
Math grades 0.404* (0.204) −0.196 (0.165) −0.279 (0.227)
Private school 0.951** (0.314) −0.156 (0.253) −0.112 (0.295)
Percent minority in the school −1.368* (0.547) −0.183 (0.577) 0.484 (0.674)
Constant 1.751*** (0.380) 0.094 (0.383) −1.403* (0.583)
N
Students 97 94 71
Schools 16 16 14
Open in a new tab

Robust standard errors in parentheses

***

p<0.001,

**

p<0.01,

*

p<0.05

The middle column in Table 2 present the results for male students who were neutral at time 1. In contrast to the results for ‘rejecters’, we do not find a significant effect of having more female peers in the classroom. Additionally, teacher gender is not a significant predictor of changes in students’ views at the end of the year. The views of these ‘middle of the road’ students do not shift in relation to their academic or social background, nor to characteristics of their schools.

The last column shows results for male students who endorsed the stereotype at the beginning of the year. Having a higher percentage of female peers does not significantly predict a change in males’ scores at the end of the year. Yet among these students there is a negative and significant effect for having a female rather than a male teacher. Thus among male students who began the school year endorsing the stereotype that males have natural superiority in STEM fields, those who had a female teacher significantly decreased their stereotypical views by the end of the year. Additionally, we note that Hispanic students and those with higher levels of parental education level were likely to increase their stereotypical views over the course of the year, all else equal.

Results for Female Students

Moving to the results for female students (see Table 3), the results reveal that among those who rejected the stereotype at time 1 (which is more than 75 percent of females in the sample), neither having more female peers in the classroom nor a female teacher significantly predicts a change in their stereotypical views at the end of the year. Put differently, while all of the females in this group initially rejected the stereotype that males are better at STEM at the beginning of the year, they do not significantly change their level of their disagreement after exposure to counter-stereotypic evidence in the form of female teachers and peers. The only significant effects we observe are for race/ethnicity and math grades. Specifically, Hispanic girls are more likely than non-Hispanic white girls to increase their stereotypical views, and those girls who report making A’s in math are more likely than those with lower grades to increase their stereotypical views over the year.

Table 3.

Predicting Change in Gender/STEM Stereotypes for Female Students by their Initial Level of Endorsement

Rejecters at Time 1 Neutral at Time 1
Gendered Environment of Classroom
Female peers (percent) 0.326 (0.550) −1.506 (1.328)
Female teacher 0.262 (0.275) 0.638 (0.584)
Race/Ethnicity (reference: non-Hispanic white)
Hispanic 0.452* (0.223) 0.209 (0.544)
Asian 0.25 (0.217) −0.742 (0.533)
Other race/ethnicity 0.14 (0.354)
Parent education level −0.044 (0.093) 0.342 (0.235)
Math grades 0.443** (0.173) −0.379 (0.364)
Private school −0.418 (0.297) −1.369* (0.576)
Percent minority in the school −0.174 (0.517) −1.348 (1.023)
Constant −0.138 (0.447) −0.383 (0.760)
N
Students 70 25
Schools 15 8
Open in a new tab

Robust standard errors in parentheses

***

p<0.001,

**

p<0.01,

*

p<0.05

The second column in Table 3 shows result for female students who had neutral views at the beginning of the year (recall that no females fell into the endorse category). Although this group is very small (n=25) we include the results for the reader’s reference. Again, we see no significant effect of either teacher gender or the percent female in the classroom, although girls attending private school become less stereotypical over the course of the year. Overall, as anticipated, there is little to note regarding the results for female students. They overwhelmingly reject the male/STEM stereotype at the beginning of the year, and having more female peers or a female teacher does not change this.

Sensitivity analyses

To ensure that the effects of female peers and female teachers were significantly different for male students as displayed in Table 2 and female students in Table 3, we utilized Stata’s post-estimation command (suest) to test differences in coefficients. Results confirmed that the effect of female peers on decreasing stereotypes for male ‘rejecters’ was significantly different than the effect for both groups of females, as well as the other two male groups. Additionally the negative effect of having a female teacher (or the positive effect on reducing stereotypes) observed among male ‘endorsers’ was also significantly different compared to all other groups. Thus the effect of female peers appears to be unique to those male students who already discounted gender stereotypes, while the effect of a female teacher is unique to those males who endorsed such stereotypes.

To ensure that our results are not sensitive to where we placed the thresholds to distinguish between those who rejected or endorsed gender stereotypes at time 1 and those who were neutral, we examined different cut-points. Specifically we lowered the threshold to be in the ‘rejecter’ category to 2, expanded the neutral category to include those scoring between 2 and 4, and restricted ‘endorsers’ to those who scored more than 4 at time 1. Our results regarding peer and teacher gender were un-changed. Additionally, we ran analyses omitting those students who initially scored a 1, as they could not become less stereotypical, as well as omitting those who scored a 5, as they could not become more stereotypical. Adjusting for these floor and ceiling effects did not change any of our results. We also note that in place of change scores, we ran lagged models predicting time 2 views while controlling on time 1, which yielded results very similar in size and significance to those shown here. Finally, we also tested an interaction between teacher gender and percent female in the classroom, to test whether the combination of these factors might have an impact on students’ views; the interaction was never significant and did not improve model fit.

Discussion

In an effort to increase the presence of women in STEM fields where they are severely under-represented, such as engineering, scholars and practitioners have generally focused their efforts on understanding what deters women and what obstacles they face, and subsequently working to help women overcome these impediments. Comparatively little attention is paid to the role of their male peers in creating and maintaining this segregated system. Indeed, recent research on programs for undergraduate women in science and engineering revealed that although program directors were very aware of negative classroom climates and peer relationships as factors discouraging women’s entry and persistence in STEM fields, their programs nevertheless focused on giving women as individuals different resources and support rather than confronting or pressing for change in the surrounding structure or environment (Fox, Sonnert, & Nikiforova, 2011). Yet as these authors acknowledge, change will not happen until the environment changes and becomes more inviting and less hostile. Our study sheds light on factors that could help to make a more welcoming environment for future generations of women by changing the views of their male peers who will be their future classmates and coworkers.

Specifically, drawing on previous research on the importance of counter-stereotypical cues for encouraging girls in STEM fields, as well as the literature on intergroup contact theory, we considered whether the gender composition of students in a high school engineering classroom as well as the gender of the engineering teacher might change boys’ stereotypical views. Importantly, we considered how this varied based on students’ stereotypical disposition when they entered the classroom, as their receptivity to intergroup contact and to re-evaluating their views towards gender is likely influenced by how entrenched or firm their current beliefs are. While we found that on average boys did not endorse these gender/STEM stereotypes, there was important variation in the strength of initial beliefs and how they changed (or did not change) in relation to the gendered environment of the classroom. Our results found that boys who initially scored low on the stereotype scale, or ‘rejecters’, become even less stereotypical in their viewpoints over the course of the year when exposed to more female peers. Although we cannot test the explicit mechanism for why this occurred, we suggest that boys who were already low on stereotypical beliefs may have been more likely to engage with their female peers in the class, providing them even more information to further dissuade them from stereotypes.

In contrast, we found that boys who already held strong stereotypical beliefs, or ‘endorsers’ were not significantly influenced by the presence of more female peers, but instead became less stereotypical in their beliefs when they had a female teacher. Perhaps for these boys, the added weight of the authority of the teacher was needed. Psychological research provides some support for this, noting a link between authoritarian viewpoints and prejudicial beliefs in general, as well as the endorsement of traditional gender roles. Specifically, those who hold strong beliefs about the importance of hierarchy and acknowledging authority are more likely to view the world in absolute terms, leading to biased views about differences between groups, and a tendency to endorse traditional gender values that articulate very different roles for men and women (Duncan, Peterson, & Winter, 1997; Peterson & Zurbriggen, 2010). Following this logic, it is plausible that boys who held strong explicit beliefs about male superiority in STEM may also be more authoritarian in their world view; thus while they might ignore the contributions of female peers in their classroom, the expertise of the teacher might be harder to discount. Thus when confronted with a female in an authority position in a STEM environment, their assurance in their ‘natural’ ability and females’ inferiority could be substantially weakened. In addition to advancing prior research on intergroup contact theory by providing empirical evidence of the need to consider differences in individuals’ initial views (Pettigrew, 2008), our study also highlights the need to consider why certain gender cues in the local environment may be the most powerful for changing minds and behaviors (Perry & Pauletti, 2011).

The results of our study have potential implications for the current debate about single-sex schools and classrooms. Within the public and political conversations on this topic, advocates of single- sex reform often suggest that girls in female-only environments may be more inclined to pursue math and science when boys are not around to discourage them, and that boys would benefit both from the absence of ‘distracting’ females peers and the presence of male teachers as role models. While the actual empirical literature on the benefits of single-sex learning is very mixed, we also note that it focuses almost exclusively on girls (Pahlke, Hyde, & Allison, 2014). Yet our results suggest that a key lever to reducing the stereotypical views of male students may be the greater presence of females in the classroom (or alternatively, a lower presence of males) either as peers or as teachers. Therefore, future research should consider the potential reifying effects of all-male STEM learning environments on young males’ views regarding females’ ability, as such contexts could serve to strengthen or at least leave unchallenged boys’ stereotypic beliefs. Furthermore, we suggest that when single-sex education advocates propose that the solution to girls’ relative lack of confidence or interest in STEM fields is to remove the boys, this sends the problematic message that boys themselves do not need to change in order for progress towards gender equity to occur.

As with any study, ours has limitations. We do not have data to allow us to examine the specific kinds or frequency of interaction that occurred between male students and their female peers and teachers. Future studies that consider the benefits of intergroup contact for gender equity would benefit from observational data, as well as interviews asking respondents to reflect on their experiences interacting with the other gender. Nor do we have data that show whether and how boys’ change in beliefs impacted the classroom environment and/or girls’ experiences. Additionally, while our study utilized measures of explicit bias, measures of implicit bias are better able to ascertain levels of unconscious bias that may be held even by those that consciously disagree with the stereotype of males’ innate superiority, including females (Cvencek et al., 2015). Indeed, we expect that our measure is likely a conservative estimate of the strength of stereotypical views held by many male (and even female) students.

These limitations notwithstanding, our results speak to the malleability of males’ views in response to gendered cues in the classroom, and thus suggest a promising avenue for more research in this area. Future studies could consider how a more project-based curriculum that promoted interaction between male and female STEM majors in college could have similar benefits, as well as whether and how interaction with female peers and female teachers (or other female experts) in certain contexts could perhaps even prevent young boys from ever developing such stereotypical views.

Conclusion

It is clear from a large body of research that discrimination and bias, as well as more subtle forms of discouragement, continue to exist in STEM majors and occupations (Clancy et al., 2014; Ecklund et al., 2012). Additionally, those individuals who endorse gender stereotypes are likely to engage in or contribute to this problem (Bobbitt-Zeher, 2011; Reuben, Sapienza, & Zingales, 2014). Rather than simply advising the relatively few women who are present in such fields to ‘lean in’ and overcome such issues, systematic change is only likely to occur if the predominantly male majority changes their views and actions. Our study highlights the importance of this issue, and contributes to the discussion of how that might happen.

Acknowledgments

Funding Sources:

This research was supported by grant (5 R24 HD042849, Population Research Center) awarded to the Population Research Center at The University of Texas at Austin by the Eunice Kennedy Shriver National Institute of Health and Child Development, and also by a grant from the National Science Foundation (DUE-0831811; Dave Allen, PI). Opinions reflect those of the authors and do not necessarily reflect those of the granting agencies.

Contributor Information

Catherine Riegle-Crumb, University of Texas at Austin.

Chelsea Moore, University of Massachusetts, Amherst.

Jenny Buontempo, University of Texas at Austin.

Works Cited

  1. Allport Gordon W. The nature of prejudice. Reading, MA: Addison-Wesley; 1954. [Google Scholar]
  2. Asbrock F, Gutenbrunner L, Wagner U. Unwilling, but not unaffected—Imagined contact effects for authoritarians and social dominators. European Journal of Social Psychology. 2013;43(5):404–412. [Google Scholar]
  3. Bench SW, Lench HC, Liew J, Miner K, Flores SA. Gender gaps in overestimation of math performance. Sex Roles. 2015;72:536–546. [Google Scholar]
  4. Bettinger EP, Long BT. Do faculty serve as role models? The impact of instructor gender on female students. American Economic Review. 2005:152–157. [Google Scholar]
  5. Blanton H, Crocker J, Miller DT. The effects of in-group versus out-group social comparison on self-esteem in the context of a negative stereotype. Journal of Experimental Social Psychology. 2000;36(5):519–530. [Google Scholar]
  6. Blanton H, Christie C, Dye M. Social identity versus reference frame comparisons: The moderating role of stereotype endorsement. Journal of Experimental Social Psychology. 2002;38(3):253–267. [Google Scholar]
  7. Bleeker MM, Jacobs JE. Achievement in math and science: Do mothers’ beliefs matter 12 years later? Journal of Educational Psychology. 2004;96(1):97–109. [Google Scholar]
  8. Bobbitt-Zeher D. Gender discrimination at work connecting gender stereotypes, institutional policies, and gender composition of workplace. Gender & Society. 2011;25(6):764–786. [Google Scholar]
  9. Bonnot V, Croizet JC. Stereotype internalization and women’s math performance: The role of interference in working memory. Journal of Experimental Social Psychology. 2007;43(6):857–866. [Google Scholar]
  10. Charles M, Bradley K. Equal but separate? A cross-national study of sex segregation in higher education. American Sociological Review. 2002;67:573–599. [Google Scholar]
  11. Clancy KB, Nelson RG, Rutherford JN, Hinde K. Survey of Academic Field Experiences (SAFE): Trainees report harassment and assault. PLOS One. 2014;9:e102172. doi: 10.1371/journal.pone.0102172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Correll SJ. Gender and the career choice process: The role of biased self-assessments. American Journal of Sociology. 2001;106(6):1691–1730. [Google Scholar]
  13. Cvencek D, Meltzoff AN, Greenwald AG. Math–gender stereotypes in elementary school children. Child Development. 2011;82(3):766–779. doi: 10.1111/j.1467-8624.2010.01529.x. [DOI] [PubMed] [Google Scholar]
  14. Cvencek D, Nasir NIS, O’Connor K, Wischnia S, Meltzoff AN. The Development of Math–Race Stereotypes:“They Say Chinese People Are the Best at Math”. Journal of Research on Adolescence. 2015;25(4):630–637. [Google Scholar]
  15. Dasgupta N, Asgari S. Seeing is believing: Exposure to counterstereotypic women leaders and its effect on the malleability of automatic gender stereotyping. Journal of Experimental Social Psychology. 2004;40(5):642–658. [Google Scholar]
  16. Dasgupta N, Scircle MM, Hunsinger M. Female peers in small work groups enhance women’s motivation, verbal participation, and career aspirations in engineering. Proceedings of the National Academy of Sciences. 2015;112(16):4988–4993. doi: 10.1073/pnas.1422822112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. DiPrete TA, Buchmann C. The rise of women: The female advantage in education and what it means for American schooling. New York, NY: Russell Sage Foundation; 2013. [Google Scholar]
  18. Duncan L, Peterson B, Winter D. Authoritarianism and gender roles: Toward a psychological analysis of hegemonic relationships. Personality and Social Psychology Bulletin. 1997;23(1):41–49. [Google Scholar]
  19. Eccles JS. Understanding women’s educational and occupational choices. Psychology of women quarterly. 1994;18(4):585–609. [Google Scholar]
  20. Eccles JS. Gendered Socialization of STEM Interests in the Family. International Journal of Gender, Science and Technology. 2015;7(2):116–132. [Google Scholar]
  21. Eccles JS, Harold RD. Gender differences in sport involvement: Applying the Eccles’ expectancy-value model. Journal of Applied Sport Psychology. 1991;3(1):7–35. [Google Scholar]
  22. Eccles JS, Roeser R. Schools as developmental contexts during adolescence. Journal of Research on Adolescence. 2011;21(10):225–241. [Google Scholar]
  23. Ecklund EH, Lincoln AE, Tansey C. Gender segregation in elite academic science. Gender & Society. 2012;26(5):693–717. [Google Scholar]
  24. Evans MA, Whigman M, Wang MC. The effect of a role model project upon the attitudes of ninth-grade science students. Journal of Research in Science Teaching. 1995;32(2):195–204. [Google Scholar]
  25. England P. The gender revolution uneven and stalled. Gender & Society. 2010;24(2):149–166. [Google Scholar]
  26. Flynn FJ. Having an open mind: the impact of openness to experience on interracial attitudes and impression formation. Journal of Personality and Social Psychology. 2005;88(5):816–826. doi: 10.1037/0022-3514.88.5.816. [DOI] [PubMed] [Google Scholar]
  27. Fox MF, Sonnert G, Nikiforova I. Programs for Undergraduate Women in Science and Engineering Issues, Problems, and Solutions. Gender & Society. 2011;25(5):589–615. [Google Scholar]
  28. Gerber TP, Cheung SY. Horizontal stratification in postsecondary education: forms, explanations, and implications. Annual Review of Sociology. 2008;34:299–318. [Google Scholar]
  29. Gresky DM, Ten Eyck LL, Lord CG, McIntyre RB. Effects of salient multiple identities on women’s performance under mathematics stereotype threat. Sex Roles. 2005;53(9–10):703–716. [Google Scholar]
  30. Hewstone M, Swart H. Fifty-odd years of inter-group contact: From hypothesis to integrated theory. British Journal of Social Psychology. 2011;50(3):374–386. doi: 10.1111/j.2044-8309.2011.02047.x. [DOI] [PubMed] [Google Scholar]
  31. Kiefer AK, Sekaquaptewa D. Implicit stereotypes, gender identification, and math-related outcomes a prospective study of female college students. Psychological Science. 2007;18(1):13–18. doi: 10.1111/j.1467-9280.2007.01841.x. [DOI] [PubMed] [Google Scholar]
  32. Lane KA, Goh JX, Driver-Linn E. Implicit science stereotypes mediate the relationship between gender and academic participation. Sex Roles. 2012;66(3–4):220–234. [Google Scholar]
  33. Lenton AP, Bruder M, Sedikides C. A meta-analysis on the malleability of automatic gender stereotypes. Psychology of Women Quarterly. 2009;33(2):183–196. [Google Scholar]
  34. Lesko AC, Corpus JH. Discounting the difficult: How high math-identified women respond to stereotype threat. Sex Roles. 2006;54(1–2):113–125. [Google Scholar]
  35. Marshall JA, Berland LK. Developing a Vision of Pre-College Engineering Education. Journal of Pre-College Engineering Education Research (J-PEER) 2012;2(2):5. [Google Scholar]
  36. Morris LK, Daniel LG. Perceptions of a chilly climate: Differences in traditional and non-traditional majors for women. Research in Higher Education. 2008;49(3):256–273. [Google Scholar]
  37. Mullen AL. Gender, social background, and the choice of college major in a liberal arts context. Gender & Society. 2014;28(2):289–312. [Google Scholar]
  38. National Academy of Engineering. Committee on Public Understanding of Engineering Messages. Changing the conversation: messages for improving public understanding of engineering. National Academies Press; 2008. [Google Scholar]
  39. Nosek BA, Banaji MR, Greenwald AG. Math= male, me= female, therefore math≠ me. Journal of personality and social psychology. 2002;83(1):44–59. [PubMed] [Google Scholar]
  40. Oakes J. Opportunities, achievement, and choice: Women and minority students in science and mathematics. Review of Research in Education. 1990:153–222. [Google Scholar]
  41. Oswald DL. Gender stereotypes and women’s reports of liking and ability in traditionally masculine and feminine occupations. Psychology of Women Quarterly. 2008;32(2):196–203. [Google Scholar]
  42. Niederle M, Verterkund L. Explaining the gender gap in math test scores: The role of competition. Journal of Economic Perspectives. 2010;24(2):129–144. [Google Scholar]
  43. Pahlke E, Hyde JS, Allison CM. The effects of single-sex compared with coeducational schooling on students’ performance and attitudes: A meta-analysis. Psychological bulletin. 2014;140(4):1042–1072. doi: 10.1037/a0035740. [DOI] [PubMed] [Google Scholar]
  44. Pawley AL. Universalized narratives: Patterns in how faculty members define “engineering”. Journal of Engineering Education. 2009;98(4):309–319. [Google Scholar]
  45. Peterson BE, Zurbriggen EL. Gender, sexuality, and the authoritarian personality. Journal of personality. 2010;78(6):1801–1826. doi: 10.1111/j.1467-6494.2010.00670.x. [DOI] [PubMed] [Google Scholar]
  46. Perry DG, Pauletti RE. Gender and adolescent development. Journal of Research on Adolescence. 2011;21(1):61–74. [Google Scholar]
  47. Pettigrew TF. Future directions for intergroup contact theory and research. International Journal of Intercultural Relations. 2008;32(3):187–199. [Google Scholar]
  48. Pettigrew TF. Intergroup contact theory. Annual review of psychology. 1998;49(1):65–85. doi: 10.1146/annurev.psych.49.1.65. [DOI] [PubMed] [Google Scholar]
  49. Pettigrew TF, Tropp LR. A meta-analytic test of intergroup contact theory. Journal of personality and social psychology. 2006;90(5):751–783. doi: 10.1037/0022-3514.90.5.751. [DOI] [PubMed] [Google Scholar]
  50. Reskin BF. The proximate causes of employment discrimination. Contemporary Sociology. 2000;29:319–328. [Google Scholar]
  51. Reuben E, Sapienza P, Zingales L. How stereotypes impair women’s careers in science. Proceedings of the National Academy of Sciences. 2014;111(12):4403–4408. doi: 10.1073/pnas.1314788111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Ridgeway CL. Interaction and the conservation of gender inequality: Considering employment. American Sociological Review. 1997;62:218–235. [Google Scholar]
  53. Ridgeway CL, Correll SJ. Unpacking the gender system a theoretical perspective on gender beliefs and social relations. Gender & society. 2004;18(4):510–531. [Google Scholar]
  54. Robnett RD, Leaper C. Friendship groups, personal motivation, and gender in relation to high school students’ STEM career interest. Journal of Research on Adolescence. 2013;23(4):652–664. [Google Scholar]
  55. Hall RM, Sandler BR. The Classroom Climate: A Chilly One for Women? Washington, DC: Association of American Colleges; 1982. [Google Scholar]
  56. Schmader T, Johns M, Barquissau M. The costs of accepting gender differences: The role of stereotype endorsement in women’s experience in the math domain. Sex Roles. 2004;50(11–12):835–850. [Google Scholar]
  57. Schmidt WH, editor. Characterizing pedagogical flow: An investigation of mathematics and science teaching in six countries. New York, NY: Kluwer Academic Publishers; 2007. [Google Scholar]
  58. Shauman KA. Occupational sex segregation and the earnings of occupations: What causes the link among college-educated workers? Social Science Research. 2006;35(3):577–619. [Google Scholar]
  59. Seymour E, Hewitt NM. Talking About Leaving: Why Undergraduates Leave the Sciences. Boulder, CO: Westview Press; 1997. [Google Scholar]
  60. Smeding A. Women in science, technology, engineering, and mathematics (STEM): An investigation of their implicit gender stereotypes and stereotypes’ connectedness to math performance. Sex roles. 2012;67(11–12):617–629. [Google Scholar]
  61. Smith JL. The interplay among stereotypes, performance-avoidance goals, and women’s math performance expectations. Sex Roles. 2006;54(3–4):287–296. [Google Scholar]
  62. Smith SJ, Axelton AM, Saucier DA. The effects of contact on sexual prejudice: A meta-analysis. Sex Roles. 2009;61(3–4):178–191. [Google Scholar]
  63. Snyder TD, Dillow SA. Digest of Education Statistics 2011 (NCES 2012-001) National Center for Education Statistics, Institute of Education Statistics, US Department of Education; Washington, DC: 2012. [Google Scholar]
  64. Stout JG, Dasgupta N, Hunsinger M, McManus MA. STEMing the tide: using ingroup experts to inoculate women’s self-concept in science, technology, engineering, and mathematics (STEM) Journal of personality and social psychology. 2011;100(2):255. doi: 10.1037/a0021385. [DOI] [PubMed] [Google Scholar]
  65. Tai Robert H. An examination of the research literature on project lead the way. 2012 Retrieved from http://www.PLTW.org.
  66. U.S. Department of Education, National Center for Education Statistics (NCES) Digest of education statistics 2012-001. 2012. [Google Scholar]
  67. Xie Y, Shauman KA, Shauman KA. Women in science: Career processes and outcomes. 73.4. Vol. 26. Cambridge, MA: Harvard University Press; 2003. [Google Scholar]
  68. Young DM, Rudman LA, Buettner HM, McLean MC. The influence of female role models on women’s implicit science cognitions. Psychology of Women Quarterly. 2013;37(3):283–292. [Google Scholar]

RESOURCES