How Stereotypes and Relationships Influence Women and Underrepresented Minority Students' Fit in Engineering

Women and underrepresented minority (URM) undergraduates declare and complete science, technology, engineering, and mathematics majors at different rates in comparison to majority groups. Explanations of these differences have long been deficit oriented, focusing on aptitude or similar characteristics, but more recent work focuses on institutional contexts, such as academic climate and feelings of belonging (fit). This study examines the experiences of women and URM students in engineering undergraduate programs, focusing on how they fit, experiential factors affecting fit, and how fit is mitigated by social relationships from their networks and organizations in which they participate (i.e., social capital). Thematic analysis of 55 women and URM interviewee responses shows that students who fit well were those with majority characteristics, including race (i.e., White, White‐passing) and gender (i.e., men, masculine appearance), and those in groups well represented in their programs numerically (i.e., men, Asian). In contrast, women and Black students encountered threats to their fit due to stereotyping from bias and differential treatment from others (i.e., being excluded from group work). However, students received advice from their social networks (i.e., family, professors) in which they were warned to expect discrimination, or through organizations in which they participated (i.e., National Association of Black Engineers) where their sense of community was expanded. The advice and resources provided through this network‐based and participatory social capital mitigated fit for women and Black students, albeit in different ways, helping to preserve their feelings of belonging and promote their persistence in engineering. We offer suggestions to enact university policies to increase access to social capital with homophilious alters and educational opportunities for majority groups.

Keywords: diversity, equity, and inclusion (DEI); feelings of belonging in STEM; network‐based social capital; participatory social capital; social networks

Extant research reveals many challenges to creating diverse and inclusive environments within US science, technology, engineering, and mathematics (STEM) educational contexts (McGee, 2020; Morton & Parsons, 2018; National Science Foundation, 2019). Women and underrepresented minority (URM) students, whom the National Science Foundation (2019) notes includes African American, Latinx, and American Indian students, are less likely to declare a STEM major in college, and URM students are also less likely to complete such a program within 5 years when compared to White men and Asian students (Fletcher et al., 2021; Mau, 2016; Riegle‐Crumb et al., 2019). For much of the past 50 years, the pipeline model was used to explain numeric declines in the overall STEM workforce before graduation or job entry. The "leaky pipeline" was often used as a metaphor for attrition, particularly that of women and URM students (Metcalf, 2010). Numerous studies have investigated women and URM students' characteristics (e.g., Cech et al., 2011; Faulkner, 2009; Goldrick‐Rab, 2006) and experiences (e.g., Beasley & Fisher, 2012; Chang et al., 2014; Espinosa, 2011; Foor & Walden, 2009; Gregory, 2015; Gregory, 2016) to address STEM persistence.

However, while academic preparation and socioeconomic status (as structural issues) are contributing factors, the lower numbers of women and URM students in STEM are due to factors unrelated to individual aptitude and interest (Acker & Feuerverger, 1996; Hill et al., 2010; Margolis et al., 2000; McGee, 2020; Sax, 2008; Seymour, 1999; Shapiro & Sax, 2011; Summers & Hrabowski III, 2006). For example, Riegle‐Crumb et al. (2019) showed that Latinx and Black students are more likely to switch from STEM to non‐STEM majors, but when controlling for social background and institution type, this trend for Latinx students disappears. Further control for high school academic preparation does not change this increased likelihood of STEM departure for Black students (Riegle‐Crumb et al., 2019). Likewise, Seymour and Hewitt (1997) revealed that women and URM students' preparation and aptitude did not account for their departure from STEM; instead, students left engineering programs because they felt disappointed with the field, were no longer interested in engineering, or lost confidence in their ability to do well in the field. In a revisiting of that landmark study, Weston (2019) found that students who left STEM cited not fitting in socially in their initial STEM major department or being unable to find help to address problems they encountered as reasons for their departure. This focus on program culture as a cause of attrition marks a turn from the previous decades of focus on the effect of students' individual efforts on attrition.

Subsequent research has focused on program culture and the people in students' social networks who help them navigate it. For example, scholars have demonstrated how STEM culture privileges experiences of white wealthy men (Parsons, 1997), contributing to the underrepresentation of Black students in STEM (Riegle‐Crumb et al., 2019). Studies have also confirmed that students link their disappointment with the discipline to inadequate advice to address problems from people in their engineering program in roles specifically responsible for providing such advice, including engineering professors (Marra et al., 2012; Nyquist et al., 1999) and advisors (Prieto et al., 2009; Schmidt et al., 2012; Sutton & Sankar, 2011). Yet, investigations have also highlighted the social capital Black students mobilize to fit in and persist in STEM in the face of these structures (Morton & Parsons, 2018).

Although differential program experiences and (lack of) assistance in addressing related obstacles arising are important structural factors affecting STEM persistence, it is not well understood how receiving assistance from others might mitigate the effects of the differential engineering program experiences on students from various groups. Most research does not often combine original inquiry into both topics. As an exception, scholarship in mathematics education confirms the relevance of structural programmatic obstacles as well as same‐gender and same‐race (i.e., homophilious) sources of guidance and self‐affirmation in shaping persistence, particularly for Black women (Borum & Walker, 2011; Brown & Josephs, 1999; Evans et al., 2011; Joseph et al., 2017; Leyva et al., 2021; Morganson et al., 2010). Yet, this work predominantly focused instructional issues (Leyva et al., 2021) or other students (Morton et al., 2019), informing less on the more social aspects involving students at they move through their STEM programs and engage with a range of other people (e.g., peers, professors, staff, industry members). Questions remain on the more microlevel processes of this social capital application, including those on how these processes may play out in engineering, especially when it comes to how students feel they belong in light of the traditionally unwelcoming STEM culture writ large.

To address this, we examine the concept of student "fit" or feelings of belonging or being welcomed into their undergraduate engineering programs. We view fit as a systemic issue reflecting program culture, and not individual students' perceived strengths or weaknesses. We situate fit within the theory of cultural models of education—how people think about and understand schooling and specifically the people part of the education process. In this study, which focuses on the engineering context, we position fit as an element of a specific type of cultural models of education, referred to as "cultural models of engineering success" (CMES) (Smith et al., 2015). Given the relationship between fit and STEM degree program persistence, we examine the effect of advice from influential people in a student's network (i.e., social capital) on student feelings of fit. Our research questions focus on (1) how women and URM students feel they fit in engineering, (2) how the experiences of women and URM students within their engineering programs shape their feelings of fit, and (3) how people with whom they interact with in their social networks and organizations influence their feelings of fit.

Our work addresses the need to better understand how social capital affects fit to extend research and inform interventions universities can enact to improve engineering programs and persistence. Given the valid critiques of the past emphasis on individuals in attempts to explain engineering persistence writ large and the more recent and appropriate focus on the structural causes of differences in STEM graduation rates, we employ structural frames to examine fit and social capital. Persistence rates in engineering are structural; there are patterns in who leaves and who stays, and these patterns are not caused by individual weaknesses in people from groups who more often leave. Instead, we assume that structural factors, and we focus on those at the group interaction level, are a primary mechanism driving patterned departure differences. Structural theoretical frames allow us to center how patterns in women and URM students' experiences resulting from how others treat them comprise the structural factors affecting their fit. This is different from previous approaches which investigate race/ethnicity or gender as causing poor fit in that we locate the issue in the mistreatment, bias, and stereotyping against these groups predominantly by people with majoritized race/ethnicity and gender identities. To underpin our work, we mobilize cultural model theory to look at fit as a cognitive outcome of relational structures by people involved in the schooling process, and social capital theory to look at how people engage with others in webs or structures of relationships in which they gain access to resources that affect fit.

Cultural model theory examines how individuals cognitively organize and structure their beliefs about material phenomena, experiences, behaviors, and the environment in specific cultural domains such as education, gender, and ethnicity (D'Andrade, 1995; Shore, 1996). Cultural models are internalized yet malleable, mediating devices that are shaped by cultural experiences, parents, teachers, and peers, but are also influenced by the wider society (Strauss & Quinn, 1997). Cultural models of education refer to the ideologies (ideas and related practices) in schooling, including assumptions about students, teachers, and the schooling process (Fryberg et al., 2013; Fryberg & Markus, 2007). Specifically, cultural models of education include beliefs about (1) how to become a successful student, (2) how the teacher‐student relationship should function, and (3) how education contributes to making an individual a better person (Smith et al., 2015).

Cultural models are distinguished from mere general knowledge in that the beliefs, meanings, and assumptions about the cultural domain are shared to a certain degree among individuals in a cultural group (Fryberg & Markus, 2007; Holland & Eisenhart, 1990; Mukhopadhyay, 2004). Cultural models of education often vary by gender and ethnicity (Fryberg et al., 2013; Fryberg & Markus, 2007). The set of beliefs about how students can succeed that are reproduced in engineering programs may conflict with those that women and URM students bring (Goldrick‐Rab, 2006). STEM and its subfields likely have overarching and individual nuances in cultural models of education (Brown et al., 2005). Our work is situated within CMES—how students understand and conceptualize engineering education, including the people involved (Fryberg et al., 2013; Fryberg & Markus, 2007; Smith et al., 2015).

We argue that fit, feelings of belonging or inclusion, is a crucial element nestled within students' cultural models of education, including their CMES. "Fit" refers to how students feel welcomed in their engineering program. Thus, fit is a characteristic of the culture and structure of the university and engineering department, not an individual characteristic based on a student's supposed abilities. Further, fit can differ within an engineering program, for instance, in zones such as in lab spaces or department offices (McGee, 2020) and across intersectional identity groups; it is affected by microaggressions against people with marginalized identities or the absence of experiencing such microaggressions.

Social capital theory refers to the repository of connections to people (alters), as well as the resources available through those connections, that aid individuals in achieving goals they might have otherwise been unable to accomplish on their own (Lin, 2001; Lin, 2008; Van der Gaag & Snijders, 2005; Yosso, 2005). Resources from alters can include instrumental support, such as concrete information and advice about how to pursue and engineering degree. Alters also provide expressive social capital in the form of emotional support and encouragement to keep going in the face of challenges (Puccia et al., 2021; Yosso, 2005). Social capital theory has been investigated at the group level (Bourdieu, 1986; Coleman, 1988; Putnam, 1996) as well as at the individual level (Burt, 1982; Lin, 1982; Marsden & Hurlbert, 1988; Portes & Sensenbrenner, 1993). We employ Lin's (2001) approach, which assumes that social capital is transmitted through social networks, focusing on students and the people in their networks, allowing us to examine how social capital may be deployed to affect fit.

Previous research has organized social capital by the types of relationships existing between individuals. For instance, Yosso (2005) demarcates between familial capital as the resources and knowledge accessed through family members, and social capital as the system of networks including peers and others outside of the family to which one has access. Swail and Hosford (2007) differentiated between primary (e.g., family members) and secondary (e.g., friends, school actors) alters.

We likewise differentiate between the type of capital we are focused on based on social locations, but use network‐based social capital and participatory social capital (Skvoretz et al., 2020). Network‐based social capital "emphasizes the networks of social ties in which an individual is embedded ... where members of the group gain access to resources (e.g., knowledge, materials, or privilege) available through their social networks" (Skvoretz et al., 2020, p. 594). These include high school teachers, college professors, school advisors, family members, and engineering peers. The characteristics of the alters in a student's social network may indicate the types of knowledge or resources to which they may have access in support of their pursuit of an engineering degree. Participatory social capital refers to how individuals may acquire knowledge, resources, and connections based on their participation in organizations and/or institutions that purposefully make information and resources available (Skvoretz et al., 2020). For example, students often benefit from participatory social capital based on the type of high school they attend, as schools differ in the types of resources, activities, and programs available to students. Alters met thought professional engineering organizations likewise provide to students important advice and resources influential in their engineering persistence, especially for URM students (Smith et al., 2021).

We situate our work by reviewing research related to how cultural models, including fit, and social capital can impact engineering persistence. We begin by identifying the structural factors which have been demonstrated to affect student fit. For instance, the reasons STEM students frequently provide for not fitting in include: experiencing a poor climate or mismatch with a degree program culture (Duncan & Zeng, 2005; Fisher et al., 1997; Robinson & McIlwee, 1991; Seymour, 1999; Tyson et al., 2010; Vogt et al., 2007; Wao et al., 2010)—including being a people‐person in a field that prioritizes impersonal relationships (Eastman et al., 2017), being made to feel like they do not belong (Marra et al., 2012), experiencing heightened anxiety due to the fear of being negatively stereotyped (Beasley & Fisher, 2012; Massey & Fischer, 2005; McGee, 2020), becoming familiar with the hypercompetitive environment (Seymour & Hewitt, 1997), perceiving discrimination and bias (Margolis et al., 2000; National Academy of Sciences, 2007), or feeling isolated (Blosser, 2020; Daily et al., 2007; Dancy et al., 2020; McGee, 2020; Rainey et al., 2018). Below we report on research that addresses how, as a result of STEM program cultures founded upon experiences of white and upper class men (Parsons, 1997) women and members of URM groups often have a different sense of fit in comparison to majority groups' fit (Rohde et al., 2019). Here, we also focus on how students might mobilize social capital to deal with poor fit, noting that existing research lacks specificity in identifying exactly how social capital might mitigate fit.

Students who "fit" or feel accepted and comfortable in engineering (Meyer & Marx, 2016), chemistry (Fink et al., 2020), STEM (Rainey et al., 2018), biology (Jantzer et al., 2021), or geosciences (Pugh et al., 2021) are more likely to declare such majors and/or have lower rates of attrition in such degree programs. The cultures of STEM programs often espouse individualistic, competitive, and solitary practices, which are most often associated with members of majority racial/ethnic and gender groups (Faulkner, 2009; McGee, 2020; Morton & Parsons, 2018; Ong et al., 2018; Secules et al., 2018). White men are the most likely to fit (Rainey et al., 2018) and generally lack awareness of how race or gender impacts other students' paths in STEM programs (Dancy et al., 2020).

The ability to "fulfill the expected roles, competencies, and identity features of a successful member of their profession," known as professional role confidence, is cultivated more strongly in men than women, which increases men's likelihood of remaining in an engineering major (Cech et al., 2011, p. 642). Men's professional role confidence does not conflict with cultural ideologies that associate masculinity with technical engineering skills (i.e., the notion that cars are a man's hobby) (Faulkner, 2000). Because engineering is highly masculinized, Faulkner (2009) asserts that engineering is "gender inauthentic" for women (p. 169). The mismatch between gender and engineering identity is likely connected to the experiences and expectations established through discourse and behavior during early STEM experiences, including those in K‐12 STEM classroom settings where boys may simply tell girls what to do (Wieselmann et al., 2020). Similarly, women's professional role confidence development is challenged by their need to "negotiate[e] engineering's hegemonic culture, which often valorizes displays of masculinity (Dryburgh, 1999)" (as quoted in Cech et al., 2011, p. 647). Women must prove to others that they are skilled engineers despite gendered expectations.

STEM culture is generally anti‐Black (Bullock, 2017; Cedillo, 2018; Martin et al., 2019; Nxumalo & Gitari, 2021; Vakil & Ayers, 2019), and STEM culture creates mismatches for members of URM groups, leading to more impoverished feelings of fit (Cech et al., 2011). URM students, especially women of color, are less likely to feel they fit in STEM (Rainey et al., 2018) and they report that race and gender impact their STEM experiences (Dancy et al., 2020). For example, research analyzing interviews and journal entries of 20 undergraduates documented how women and URM students in mathematics reported stereotyping related to intelligence, with URM students reporting stereotyping related to assumptions that they engaged in "rude" behavior (Leyva et al., 2021). Stereotyping and low representation in STEM classes caused students to experience differential opportunities to participate and be supported, limited opportunities to create community with students like themselves, and feelings that only members of the majority group belonged in STEM (Leyva et al., 2021; see also Malone & Barabino, 2009). Further, Rockcliffe (2020) analyzed surveys with 645 US‐born Black women in engineering undergraduate programs, finding that when Black women engineering undergraduates experienced racial microaggressions (but not gender microaggressions) it made it more difficult for them to develop an engineering identity linked to engineering persistence due to feelings of poor fit with the profession and people in it. In response to threats to fit, Morton et al.'s (2019) analysis of interviews with 44 Black STEM undergraduates showed that students held a Black X Consciousness that shaped how they interpreted experiences and made decisions given the various social implications of what it means to be Black and in STEM. Here, Black students coalesced around other Black students, which facilitated their STEM learning (including helping them access resources) especially in light of the underrepresentation of Black students in STEM, shared STEM alienation, and strong Black identity (Morton et al., 2019). Further, Black students are also particularly affected by broader campus climates, in which anti‐Black racial microaggressions are often experienced (Lee et al., 2020). Unfortunately, poor fit inhibits students' ability to gain social capital, which could promote fit (Blosser, 2020).

A particularly relevant factor involved in women and URM students' fit in engineering is stereotype threat, which they experience at higher rates than White men (McGee, 2020). Stereotype threat is "the immediate situational threat that derives from the broad dissemination of negative stereotypes about one's group; the threat of possibly being judged and treated stereotypically, or of possibly self‐fulfilling such a stereotype" (Steele & Aronson, 1995, p. 798). Stereotype threat can result in assumptions that women or URM students are in college simply because of affirmative action policies and not because they are academically capable (Ong et al., 2018). In addition, being stereotyped negatively impacts students' anxiety, anger, imposter syndrome, and can make them want to work compulsively (McGee, 2016).

Interventions designed to improve fit by addressing stereotype type, such as through providing warnings about discrimination, have been shown to yield positive outcomes for students (Jordt et al., 2017; Walton & Cohen, 2011; Yeager et al., 2016). Yeager and colleagues quantitatively experimented with explaining to students (from all majors) that they would encounter stereotyping in college. The researchers found that preparing students for what they might encounter improved their enrollment, grade point average, and overall college experiences. The intervention increased socially and economically disadvantaged students' (including URM students') use of support services and helped them develop social networks of friends and mentors.

McGee and Martin (2011) introduced "stereotype management" to refer to the strategies that students use to cope with being stereotyped. Black students often must be "hypervigilant about negative perceptions of Black behavior" and thus engage in preemptive behaviors as they expect "to be challenged on their intellectual capacity" (McGee, 2016, pp. 1633–1634). The body of research on network‐based alters suggest but does not clearly link how they may help with stereotype management by more often providing expressive support, in addition to instrumental advice. For example, in her analysis of two rounds of interviews with 10 Black women persisting in their engineering degree programs, Gregory (2015, 2016) found that women of color in STEM manage stereotype threat with social capital by leaning on the support of family and friends and creating counterspaces—safe spaces for marginalized groups that offer a haven from microaggressions and isolation (Ong et al., 2018). Research on the influence of social connections with family or friends from professional networks on persistence found that personal relationships help students feel they fit and stick with STEM (Rainey et al., 2018), including for Latinx and African American students (McGee, 2020; Rincón & Rodriguez, 2020). Ortiz et al. (2019) identified how family members 14 Black STEM majors (half at a Predominantly White Institution [PWI] and half at an Historically Black College/University [HBCU]) provided representation of Black people in STEM or encouraged an interest and persistence in STEM.

The participatory social capital literature suggests that such alters may more often provide instrument support and indirect expressive support provided to students seeing people like themselves success. Having mentors who share common marginalized demographic identities with their mentees helps them see STEM as inclusive (Hammack & High, 2014; Kricorian et al., 2020). Professional engineering organizations, particularly ones that are gender‐focused (e.g., Society for Women Engineers) and racially/ethnically‐focused (e.g., National Society of Black Engineers, Society of Hispanic Professional Engineers) can encourage the establishment of connections with similar others as well as reduce feelings of isolation and promote persistence (Campbell‐Montalvo et al., 2021; Espinosa, 2011; Ong et al., 2018; Smith et al., 2021). Informal mentoring programs positively influence participation in research and academic progress (Schwartz, 2012). Further, marginalized students in STEM identify peer support as a support structure that is useful for them (Morgan, 2013). Course content‐related peer discussions outside of class have also been shown to increase degree persistence for women of color (Espinosa, 2011). Physical structures that place students and alters in close proximity, such as living learning communities, academic support programs, and student support centers, have also improved engineering retention (Inkelas & Soldner, 2011). Connections between faculty and students and community among peers has also been identified as important in promoting fit of Black women in STEM at HBCUs (Nguyen et al., 2021). Morganson et al.'s (2010) study of more than a 1000 undergraduate students, 40% of whom were African American, found that women more commonly used social coping provided by guidance from alters, and it helped predict their degree persistence. Ortiz et al. (2019) showed that networking opportunities such as STEM clubs that allow social activities were useful sources of building capital among Black students to pursue STEM. Morton's (2017) research with 10 Black women STEM undergraduates (half at a PWI and half at an HBCU) showed how social networks created through participating in research experiences provided various resources, including encouragement to continue in STEM, opportunities to form study groups, a community to share struggles, and access to faculty. Similarly, in their research analyzing department cultural attributes supporting the success of diverse physics students, Schipull et al. (2019) found that opportunities for students to engage in discussion of identity to negotiate marginalized statues (i.e., women, URM, LGBTQ) with STEM identity were useful. Here, coursework supporting conversations on identity and STEM, a widespread promotion of collaboration rather than an individualistic values, and a department ideology of "success together" promoted persistence. Students were encouraged to network and to discuss identity in ways that promoted fit in STEM.

Although research suggests that social capital can play an important role in students' fit and persistence in STEM and engineering, most of the studies reviewed did not explicitly identify the ways by which alters might positively influence fit or did not focus on alter mitigation in relation to stereotyping from others, nor focus on a range of alters (Espinosa, 2011; Gregory, 2015; Gregory, 2016; Inkelas & Soldner, 2011; McGee, 2016; McGee & Martin, 2011; Morgan, 2013; Nguyen et al., 2021; Ong et al., 2018; Schwartz, 2012). Likewise, these studies do not generally explain how coping nor the alters explicitly promote fit—nor is distinction often made between possible differences in how network‐based and participatory social capital function in this regard. To address this, the goal of this research is to examine the process, including the specific advice provided, through which social capital can prepare engineering students to manage stereotype threat and their fit.

To examine how social capital affects fit, we address the following research questions:

• What are women and URM engineering undergraduates' perceptions of fit in their engineering degree program?
• What experiences/interactions influence their perceptions of fit?
• In what ways do network‐based and participatory alters influence students' perceptions of fit?

We examine social capital as a potential mitigator of CMES. In this study, social capital refers to alters (i.e., people influential in a student's life) (Burt, 1982; Lin, 1999; Son & Lin, 2012) as well as the advice/resources made available through alters (Lin, 2001). We focus on people in one's network who provide advice about how to succeed in engineering or who provide support and encouragement for success (i.e., network‐based social capital) (Puccia et al., 2021; Skvoretz et al., 2020) and people one comes to know as a result of participating in organizations that focus on professional development efforts and opportunities (i.e., participatory social capital) (Skvoretz et al., 2020; Smith et al., 2021). Within participatory social capital, we focus on programs in which students can be enrolled prior to or as they begin their engineering degree, such as bridge programs, activities and design teams, and nonengineering affinity organizations. We specifically do not include professional engineering organization here given our intent to focus on the factors affecting departure at the beginning of students' program and that students are often not eligible to participate in professional engineering organizations until later on (see Smith et al., 2021 for a focus on these). Nguyen et al. (2021) found that both pre‐college and college social capital impacted STEM fit and persistence for Black women in STEM at HBCUs. To examine the interaction between social capital and fit, we focus on how initial and acquired social capital (i.e., people and accompanying resources in students' social networks prior to and after entering their engineering degree program) affect students' feelings of fit.

The participants for the study were 55 undergraduate engineering students enrolled in five universities, which included three PWIs, one Hispanic‐Serving Institution (HSI), and one HBCU. Forty‐two participants attended PWIs, eight attended an HSI, and five attended an HBCU. Notably, the campus demographics at the three PWIs differed; in 2014–2015 at two PWIs there were twice the number of Latinx students (20%, 16% respectively) than Black students (11%, 8% respectively), while in 2016–2017, the third had low levels of both Latinx and Black students (6%, 5%). Also, the engineering program at the HBCU is operated in conjunction with an engineering college at a PWI, so HBCU students actually attend all engineering classes on a PWI campus and are therefore identified in analysis as attending PWI engineering programs. We recruited participants with both higher and lower levels of social capital, based on their responses to the first of five surveys used as part of a larger project (Skvoretz et al., 2020). Then, 32 participants who agreed to be interviewed were among the group of students with lower levels of entering social capital, 22 were among those with higher levels of social capital, and one was from group of students with a middle range of social capital. Table 1 shows the gender and race/ethnicity of study sample.

1 TABLEInterview participant demographics by race/ethnicity and gender

Eligible students were those in the first year of their engineering program in the fall semester of 2014. We conducted the interviews during the 2015 calendar year, so students were at the end of the first year in their engineering programs or starting their second year.

Participants were enrolled in US engineering programs, which are usually classified as four‐year programs culminating in a Bachelor's degree, but many students may take more than 4 years to graduate (American Society for Engineering Education, 2016). The programs often require high grade‐point averages for entrance, including previous coursework in science and mathematics. The first years of college coursework often include general education courses, STEM courses, and large lecture‐based, entry‐level engineering courses that operate as a "weed‐out" mechanism in which grade curving and other logics may be used to enable a certain number of students to move forward in the program (Leyva et al., 2021; Weston, 2019). The later years of engineering programs include more advanced engineering coursework targeted toward students' majors. In the US context, colleges of engineering typically include majors such as mechanical engineering, chemical engineering, biomedical engineering, computer engineering, aerospace engineering, and the like. Each of the engineering subfields differs in the percentage of gender and ethnic/racial student demographics, for instance, chemical engineering often enrolls more women, but overall in the United States, engineering is commonly considered a "male profession."

The interview protocol was designed to measure how CMES and social capital affect students' persistence and retention in their engineering program. We used each participant's survey responses to personalize the interview protocols by asking about the alters they had indicated were influential to them and their pursuit of engineering, including mention of the alters' demographic characteristics (Skvoretz et al., 2020). Specifically, the interview guide contained items asking about participants' perceptions of whether they "fit" in their engineering program, what sort of stereotypes about engineering they experienced or heard, and how alters contributed to their feelings of fit (see Table 2).

2 TABLEInterview items measuring fit and relevant themes

Questions included "tell‐me‐more" probes and requested examples to encourage participants to give detailed responses (Bernard, 2011). The interviews were conducted by two Black women and one White woman. When feasible, participants were matched with interviewers of the same race to promote rapport given the potentially sensitive subject matter of the interviews. The 30‐ to 60‐min interviews were conducted in person or via videoconference, audio‐recorded, and transcribed verbatim. Each participant received a $25 Amazon gift card as an incentive.

A four‐person subset of the research team employed thematic analysis to identify themes in data by the extent to which they were patterned and by how well they related to the key concepts in this research (Braun & Clarke, 2006). Our approach was both deductive, in that we looked for the presence of data relating to our theoretical frames, and inductive, in that we were open to how the data related to the frames or exactly how our variables intersected with one another (Braun & Clarke, 2006). We assumed a realist stance as our interpretation is based on the assumption that understanding participants' experiences is relatively straightforward based on listening to what they say (Braun & Clarke, 2006), but that also their experiences likely differ based on their social positions and the structures in which they participate (Hewitt et al., 2012).

To code our data, we first developed a codebook including primary code and subcode relating to notions of fit from the literature as well emergent from the interviews (see Table 3).

3 TABLERelevant codes used in analysis

To reach consensus about codes and the meanings attributed to them, we engaged in several rounds of discussion that resulted from our independent reviews of interview transcripts and our collective understanding of the literature. These reviews were done with a focus on cultural models and social capital, and with the understanding, our interview protocol had been formulated to gather data in relation to those frameworks. We collectively drafted the first version of the codebook during several days of group workshop meetings wherein we discussed the data and frameworks. The codebook contained parent codes (i.e., cultural beliefs: fit, social capital) with child codes focused on level of analysis (i.e., high school, postsecondary) and themes (i.e., stereotyping, type of alter). We coded an interview together to determine how coding would proceed (i.e., interviewer questions would be included in the coded passage to provide context in analysis). This exercise required us to refine the draft codebook. Over several days, we engaged in numerous rounds of individually coding the same interviews and then reconvening to compare codes, refine the codebook, and qualify code meanings. For codes not initially clear, we included a definition or example in the codebook to promote coding fidelity and accuracy.

Once we reached a consensus about the codebook, two of us coded the interview data. We first coded the same interview to establish intercoder reliability on a line‐by‐line comparison (Bernard, 2011; O'Connor & Joffe, 2020). Our intercoder reliability was 90% when determining level of agreement for primary codes only. Our intercoder reliability was 83% when determining level of agreement for both primary and secondary codes. We deemed this level of agreement sufficient because 80% reliability is often noted as the benchmark in literature (Krippendorff, 2003; Landis & Koch, 1977). The coders each coded half of the 55 interview transcripts on paper and another team member entered the codes into QSR NVivo 11.0.

To analyze the coded excerpts, we exported passages assigned a particular code into a Microsoft Word document to review them collectively in light of our research questions and theoretical frames. We identified themes through memoing and group discussions of potential findings (Bazeley & Jackson, 2013; Braun & Clarke, 2006). Summary statements describing themed groups of data helped us structure the findings presented. Finally, we identified sample quotes from a range of participants to exemplify the identified themes.

The results show how women and URM engineering undergraduates fit in their programs based on their gender and racial/ethnic identities.[1] As shown in Table 4, women and URM students' issues with fit arose from feeling isolated or being one of the few people like themselves in their engineering program. Likewise, women and URM students with majority or majority‐adjacent identities (i.e., White women, Latinx students) (Bonilla‐Silva, 2006), or those with less "net vulnerability" (Morton & Parsons, 2018; Spencer, 2006) tended to fit better in their programs than students with fewer majority statuses (i.e., Black women). The types of experiences that influenced students' perceptions of fit related to gender and racial stereotyping. As with fit more generally, outcomes of stereotyping were related to participants' identities: gender stereotyping hurt women's fit, but it was not reported as a concern of men. Additionally, Black students, in comparison to White women and Latinx students, reported the most egregious threats to their fit in the form of racial prejudice and discrimination (i.e., stereotype threat). However, network‐based alters provided women and URM students, particularly Black students, with advice in the form of warnings that people would treat them differently because of their race and/or gender—and the advice helped students cope with sexism and racism (i.e., stereotype management). Students also benefitted from the community they received from participating in organizations where they got the chance to get to know others like themselves and feel included.

4 TABLEResearch questions and themes

1 Abbreviation: URM, underrepresented minority.

We asked interview participants how they fit in engineering based on their gender and race/ethnicity. Participants generally responded that they did or did not fit in well, less often they hedged in their responses or said they were unsure how they fit or they did not directly answer the question. Of the 55 participants, 37 (67%) students stated that they felt they fit in their engineering program, 8 (15%) reported not fitting in, 5 (9%) were unsure if they fit in or hedged in their response, and 6 (11%) did not answer. Five of the eight students who indicated they did not fit were Black. There were two main themes in interviews: (1) women and URM students experienced feelings of isolation associated with being numerical minorities; and (2) women and URM students with majority statuses reported better feelings of fit.

Women of all races/ethnicities and URM students, particularly Black men, acknowledged that being one of only a few students from their demographic group in engineering played a role in their fit (or lack thereof). For example, a White woman at a PWI shared that she did not fit in because of her gender, sharing:

I stick out more than I fit in ... My department is 90% male ... There's about six girls and everybody knows their names just because they're the only girls in the department.

Responses from Black men were similar to those provided by women: being a numerical minority impacted their perceptions about their fit. For instance, a Black man shared about his experiences attending engineering classes at a PWI. He compared being a numerical minority in high school to his experiences at university where there were few Black students in STEM:

[In] high school, I might have been the only Black kid in class ... [Going to this PWI] there might be [a] sprinkle of African American kids ... It's 100, 200 new faces you never seen and they're a totally different skin color ... [I] don't know anybody.

A Latinx man enrolled in the same engineering program corroborated the environment described above, saying that Black students comprised a small proportion of the student body. The Latinx student remarked about the low numbers of Black students in his classes when talking about how his own fit as a Latinx man in the program was good when compared to others. These comments show that participants compared themselves to the quantity of others (Cech et al., 2011; McGee, 2020), and felt like interlopers if there were not others like themselves in their program.

Differences in students' perceptions of fit were influenced by others' reactions to their identities. For example, White women and Latinx students who possess majority or near‐majority identities reported having experiences that they believed were not much different from those in the dominant group. Thus, they reported fitting in better. For example, a White woman shared that although she had thought about the role of gender in engineering, she never considered how a students' race/ethnicity might affect their experiences. Similarly, several Latinx men reported that they felt welcomed in their programs because they possessed phenotypic characteristics typically associated with White people (i.e., "Caucasian attributes," such as blue eyes) or they spoke English without a Spanish accent. For example, a Latinx man enrolled in a PWI noted, "My family's very Puerto Rican ... But I mean, I have no accent. I won't really openly say I'm Puerto Rican. I don't look Puerto Rican. I have blue eyes." When asked if he felt that he fit into engineering as somebody who is Latinx or Puerto Rican he replied, "Oh yeah. I think everybody here is very open. Especially in engineering, because they teach you so much about teamwork and inclusiveness. I guess that you don't really see any troubles there." Most Latinx students, especially Latinx men, did not indicate that being Latinx affected their feelings of fit.

We found that gender and racial/ethnic stereotypes and microaggressions were common, came from peers, instructors, and guest speakers, and generally affected students' feelings of fit. While gender stereotypes negatively impacted women's perceptions of their fit, such stereotypes were reported by men as either the "truth" or their effects were dismissed. Racial/ethnic stereotypes and microaggressions particularly affected Black students' feelings of fit and were more commonly experienced by them. As with general feelings of fit, students with majority or majority‐adjacent statuses less often reported negative interactions related to gender and race/ethnicity and tended to not acknowledge how such interactions could be exclusionary.

Students from all gender and racial/ethnic groups identified gender stereotypes suggesting that women are inferior and that engineering is unnatural for women. Examples of stereotypes shared included: women enter the "easier" sub‐fields of engineering such as chemical engineering; engineering is a male‐oriented field; women leave engineering to have children; women tend to ask for help; engineering has a "lot to do with cars" which is a man's area; and women, particularly beautiful women, are not taken seriously in engineering. In some cases, students discussed these stereotypes, but did not label them as such.

For instance, students shared examples of stereotypes that suggest that women are not as smart as men. For example, a White woman at a PWI recounted how men questioned how she was able to get into her engineering program. She noted:

I've had people in my chemistry classes look at me and ask, 'How the F did you get into this school?' I've had people tell me that the only reason that I'm here is because I'm female ... The number of times that—not just students at my university but other college students or other males— ... reiterate that you're only here because you're a girl and because people care about diversity because that's politically correct is super frustrating.

The student noted that this pushback typically comes from "White males that get the right score, [but] that didn't get into the school." She stated that their animosity causes them to "look for any difference, whether it's your race, gender, socioeconomic class," which they will then attack. Likewise, another White woman enrolled in a PWI reported that she heard peers in college, as well as high school, make negative comments about women in engineering. She shared:

I've heard some comments from men, 'Oh, female engineer' or 'What are you doing?' Especially in high school when some peers found out I was going into engineering ... [they asked,] 'What are you doing that for? You're a woman,' misogynistic comments.

A subtype of this trope targets particularly beautiful women, and assumes they have special difficulty in engineering. For example, a Latinx man at a PWI shared the following:

It's not really offensive, but if you see a woman in engineering they are (A) considered ugly; (B) married and coming back [to college]; or (C) considered super geniuses ... In my statics course there are five women; and they're not bad looking, but they're all late twenty/early thirty and married ... Calc II comes along, [the number of women is] cut in half ... Most often, if a girl in engineering is good looking, they're not going to make it.

Although most gender stereotypes in engineering came from peers, participants also reported that instructors, guest speakers, and others also made such comments. When asked about a specific time she had heard people in her engineering department make a negative stereotypical comment about who should be an engineer, a White woman at a PWI recalled an email she received from the college addressing an incident when "a professor said 'I've never had a female pass this class.'" She also identified a different occasion when a graduate teaching assistant remarked negatively about women's abilities. She said:

There was a graduate student assistant ... it was in a room full of guys with one girl and he said, 'This problem is so easy your girlfriend can do it' or something like that. We all went, 'Wow.' My female friends in engineering were all pissed, were like, 'Really?'

Another White woman enrolled in a PWI stated: "The professors are mostly great, [but] I had a sexist comment from one recently. It took me down a notch."

In addition to hearing stereotypes, participants also reported that they heard about or witnessed microaggressive acts against women. Women, especially White women, reported fitting in well despite witnessing these stereotypes and microaggressive acts (we connect this later to their social capital). When asked about their experiences, women talked about the assumptions men held about women's roles in engineering group work. Women recounted stories such as being prevented from using lab equipment, being expected to take notes in lab groups or meetings, or being asked to complete less technical aspects of projects. For example, a White woman at a PWI asserted: "I know a lot of (women) who've had male partners where the group was working with power tools and men wouldn't let them do that just because they're girls, which is not cool." Later in the interview, the student mentioned again having to deal with stereotype threat where men question women's abilities to contribute to lab work requiring the use of power tools, "[They approach group work like they are thinking,] 'Oh, this is a girl. How is she going to use power tools to make our project?'" Other women reported that men expected them to take notes or to do less technical aspects of assignments rather than engage in hands‐on work. For instance, when asked about a time that she heard people in engineering make negative comments about who should be an engineer, a Black woman at a PWI said:

A design course I took freshman year, everyone takes it. We work in groups and [my group] was predominately male because that happens all the time. You get shoved to doing secretarial stuff. [They said], 'You want to type it up for us?' [I responded], 'No, I don't, I want to contribute to the ideas.' That happened a couple times ... They don't say it, but they assume you want to do less technical parts or responsibilities. It happens a lot.

When asked who made those types of statements she responded, "White men. Usually I know the Black men and they know me. They know that's a terrible thing to say and they would be checked if they said something like that." In addition to microaggressions, restricting women's access to equipment and active engagement in group work, women less frequently discussed sexual harassment, such as being "hit on" by men in their engineering program.

Overall, men tended to minimize the presence and role of gender stereotypes in engineering. Responses provided by men indicated that they believed stereotypes about women in engineering or had once believed them. For instance, when asked if he had heard any gender stereotypes, a Latinx man at a PWI talked about how he had come to believe these stereotypes based on objective (i.e., McGee, 2020; Moore & Nash, 2021), "empirical evidence":

Everyone who has made those assumptions, it's based off empirical observation. So, we all witness it and then we all [discuss it] ... it always comes. Someone tilts their head, 'Have you ever noticed?' or 'Do you think?' And we all come to a mutual agreement, 'Yeah' ... That [discussion] also includes women, like we'd always joke about it.

When asked about hearing stereotypes and its influence on their sense of fit, men explained stereotypes, such as differences in participation rates across groups, with comments such as "not everyone has a liking" for engineering. For example, when asked how he was similar to other engineering students, a Black man enrolled in a PWI compared his gender to the genders of others. He explained that he felt comfortable because he was a man and that women did not often pursue engineering. He stated:

I'm a boy, that's one. I think every engineering student has that drive. Engineering is a hard major; everyone who is an engineer really wants to be in engineering, wants to be successful, and knows what they're getting into. I have that drive, with everyone else.

When asked to clarify what he meant by saying he was a boy in his response, the student said, "I was just making like a sarcastic comment. Because everyone knows that it's statistically mostly boys in engineering, girls don't really go for engineering."

Similarly, women reported that men were often the ones enacting microaggressions or downplaying evidence of gender stereotypes in engineering. This can be seen in the earlier case where women recalled being prevented from using tools or being expected to do nontechnical aspects of group projects by men who were usually White. Likewise, evidence that men downplayed gender stereotypes can be seen in the previous example when the Latinx student linked women's attractiveness to their intelligence, simultaneously minimizing the gravity of such beliefs by beginning his comment with "It's not really offensive..." or when the White woman recalled being told women are in engineering due to diversity initiatives. Similarly, a Latinx woman at a PWI shared about how her male peers reacted during a moment that occurred when an invited guest made a gender‐related microaggression:

We had an industry panel ... One of the guys ... ran his own engineering firm. Our ChemE advisor..., she was there moderating the panel, asking questions. That guy goes..., 'Oh yeah. When you have to explain [an engineering related matter] to your mom.' And [the ChemE advisor] goes, 'Or dad' ... He didn't realize it was a microaggression. But I picked up on it and I was like, 'What an ass' ... One of the guys I was talking to said, 'Oh, but he wasn't trying to do that.' I'm like, 'I know he wasn't trying to do it, but it's a microaggression, all the girls in the room picked up on it and were pissed.'

Not only had the male student not recognized the gendered microaggression, he dismissed it when his classmate pointed it out to him.

Black students commonly and robustly recounted incidences of racial stereotyping and microaggressive acts in engineering and on campus, including incidences that demanded they overcome the stereotypes held by majority students. Although some participants recalled instances of acts against Middle Eastern, Arab, or Muslim students, most reported incidents involved negative acts against Black students.

Most stereotypes focused on students' academic abilities as a Black person, and how students spoken of them were indicative of stereotype threat. For instance, a Black man enrolled in a PWI shared the following about other non‐Black students:

A lot of them were standoffish. I could sense how uncomfortable they felt until I spoke. Then, they realized that whatever they thought about me was not true—the stereotypes about Black people being loud, violent, or not intelligent. They're looking at you from that [perspective], so you have to break down those barriers before you can get to know them ... I don't like it, but it is what it is. I've grown up dealing with and learning it.

Similarly, a Black man attending engineering classes at a different PWI shared that he and other Black men in his program often discuss their experiences with this kind of prejudice. For example, he said that the White students at the PWI "looked at us funny." He elaborated, "It's like [there were] five of us [Black students] in a room full of thirty." When asked what he thought they were thinking, he said: "You see how I'm dressed, I'm smarter than I look. But I feel like [the White students were thinking], 'He's Black, he's got this haircut ... Is he a thug? Is he smart?' I could be wrong; it's just an awkward feeling." Another Black student said of the perceptions of others: "They'll think, 'Oh those kids are going to fail, they're not interested.' But [when] we end up getting better grades in class ... it always turns their world upside down."

Additional racial/ethnic stereotypes such as Black students are lazy or poor also impacted feelings of fit. For example, a Black woman at a PWI shared:

From the things I've heard, the jokes that aren't funny. Things that classify minorities and my people as being lazy, poor, terrorists, and those things. It's disgusting to think that people generalize a population by something that happened by one or two people.

Feelings of fit (positive or negative) for URM students were also tied to their perceptions about broader campus climate. And the climate was most negative for Black students. Almost all Black participants described campus experiences in which they had to deal with negative preconceived notions about them as Black people (e.g., that they were a threat). For instance, a Black woman at a PWI shared about a flyer that circulated on campus which encouraged White women to avoid dating Black men. She recalled:

This week there were flyers posted [and] shoved under dorm doors. It was addressed last night by the campus president, and the cultural climate on campus has been heavy. One flyer detailed why White women should not date Black men and why Black men were inferior to White men. It was terrible, I couldn't finish it. It was under my door. They were all over the [engineering building atrium], posted on social media, everywhere ... Other flyers were for Euro Americans, 'Don't be afraid of them. Be proud to be White'.... Why? They weren't inclusive things. They were inconsiderate and hurtful ... It's not a fun place to be, it's tough. I'm having a harder time figuring out where I fit. As a leader, I make sure everyone's okay and we stick together. We're a community and family.

Even though her fit is threatened, the student was still concerned with the feelings of others.

Most White women and Latinx students reported that racial stereotypes did not exist in engineering (distinct from gender stereotypes, which all participants reported). Unlike Black students, Latinx students in both HSIs and PWIs usually indicated that they felt welcomed on campus. Yet, their responses often reproduced stereotypes about students of color in engineering, particularly stereotypes about Black students. Similar to how men discounted gender stereotypes, Latinx students tended to minimize the effect of racial stereotypes. For example, as noted earlier, many Latinx students felt race was "thrown out the window". For instance, a Latinx man enrolled in a PWI said:

A lot of people in my classes are Spanish. I've seen a lot of Brazilian foreign exchange students. I don't feel like an outsider or anything like that. I feel welcomed, so I don't think about being Cuban much. I hear a lot of people speaking Spanish and it seems like English is their second language, so I don't really see much of a difference.

In addition, non‐US‐born Black students cited diversity on campus as a reason for positive feelings of inclusiveness, and generally reported different opinions than US‐born Black students. For instance, a Black woman enrolled in an HSI said:

[My university] has a lot of international students. My family is from a foreign country, Jamaica. I can relate to a lot of international students because even though I was raised in Florida, I'm Caribbean because the household is everything Caribbean. There's no American traditions ... I relate to them; we can talk about the island life or foreign life.

These comments suggest a difference in experiences for Black students based on nationality.

Participant responses showed that alters from their networks and from organizations in which they participated provided resources affecting fit. Specifically, network‐based alters warned students that others would react to their minoritized statuses (i.e., stereotype threat) of gender and race, and students reported that these warnings helped them know what to expect and how to cope (i.e., stereotype management). In addition, participatory‐based alters connected students to opportunities that helped them feel included, confident, and comfortable in their engineering program more generally. In many cases, students shared that they could identify how the social capital provided by alters positively contributed to their feelings of fit in their engineering program. They also indicated that their feelings of fit improved over time.

Students reported that they received insider knowledge from individuals in their networks (i.e., advisors, parents, engineering upperclassmen, high school teachers, and internship supervisors), which helped them navigate feelings of fit. Specifically, alters often warned students about potential issues they might encounter (i.e., being excluded from participation in various activities) related to other people's attitudes regarding gender and/or race (i.e., stereotype threat), which helped them better navigate these experiences.

Most women (regardless of race/ethnicity) said that an alter had prepared them for future experiences or helped them make sense of past experiences where they had experienced gender stereotyping or harassment and encouraged them to keep going (i.e., stereotype management). For example, a Puerto Rican woman at a PWI indicated that she drew on the advice about how women are treated in engineering given by her mother, who was an engineer. She said:

I don't get fazed by it ... because my mom was in engineering and she told me stories [about] how ... a male counterpart in her department tried to undermine her and got shut down so hard he never did it again. So, it was the attitude. I've had conversations about women in engineering ... I'm like, 'I'm going to do this' and 'I'm qualified to do this.'

Her mother's advice helped her deal with negative attitudes toward women in her engineering program. A White woman at a PWI shared similar advice she received from her middle school shop teacher with whom she remained in occasional contact. She said:

[The advice] was along the lines of being a female in engineering, 'You know you're going to be one of the only women in your class. You're a tough girl and you should stay like that. You know your stuff, make sure that people know that you know your stuff and they'll take you seriously' ... I'm known as 'the hot one' in aerospace which is ridiculous and so I just wanted to prove to people that I was not 'just a girl'.

This advice helped her deal with stereotypes requiring women to prove themselves, especially as her intelligence was doubly suspect due to her appearance. She continued that an upper‐level engineering peer echoed the advice she received from her shop teacher. She shared:

I've asked [my peer], 'Hey, everybody respects you. How do you ensure that people respect you? They go to you for questions and people don't come to me for questions.' She [said], 'Well you got to prove yourself. You have to prove yourself a bit more than guys do' ... It was nice to hear it from somebody who dealt with the same thing I did.

Women also said alters prepared them to deal with sexual harassment and sexism. For example, a White woman at a PWI talked about her discussions with a woman engineer for whom she worked as an intern. The student recalled:

The biggest non‐technical takeaway I got from her is how she manages her personal life with her professional life ... how you deal with sexual harassment from your boss. ... One of the male co‐ops ... had been there a similar amount of time with similar qualifications. I was trying to get approval for this project and every time I went through the proper channels, I talked to the right people with ample amount of time beforehand, I got stonewalled. So, we sent the male co‐op and he had no problem getting people do it.

Besides receiving academic and career advice from her mentor, she also received insider information she needed to overcome the additional barriers she faced as a woman in engineering. In the example, although she had been prepared, her colleague (a man) was able to move things along when she could not. She explained how her boss helped her make sense of it:

I was frustrated by it. But my boss looked at me, and she goes, 'You know people aren't going to come out and tell you this, but that's what sexism is really like.' Having somebody that can call it like they see it without feeling like [you were making it up is important]. Especially as a woman in engineering, you get told, 'Oh that's in your head.'

While she indicated that she was upset by the difference in treatment between her and the male co‐op, she also noted that her alter's advice helped her feel validated that what she experienced had really happened and that it was wrong. Similarly, the White woman from a PWI, who earlier shared the story about men not allowing women to use power tools, shared that such instances of differential treatment did not affect her as much as they might have affected others because an engineering peer had prepared her. She said, "One of my closest friends in engineering is a girl, so I feel like I always have her to go to [because we share] similar perspective[s]."

Similarly, Black students reported the value of advice from network‐based alters about stereotype management. For example, a Black man at a PWI shared that his parents had prepared him for others having a negative opinion of him. He shared:

I expected [that people would have prejudice related to race] because my parents would tell me, 'This is what you should expect going to a public White institution, people already have these thoughts about you and you have to show them that you're not like that.' I did what they said and everything was good.

In the end, he felt everything worked out because he followed his parents' advice.

Black women faced the double bind of having two marginalized identities. For instance, a Black woman enrolled in a PWI recalled that her college advisor said she should be ready to show that Black women are not failures when her peers say they think she is lazy or when professors ask if she is in the right classroom. When asked about a specific time when people in her engineering department made comments about who should be an engineer, she answered:

It was in one of my classes. Students didn't want to be in a group with me because they thought I'd be lazy. I ended up not having a group and my professor had to assign a group for me out of the leftovers. It was funny that the other people in the group weren't lazy or other stereotypes. We got one of the best grades on our project and people wanted me to be in [their] group after and I was like, 'I don't want to be in a group with you.'

When asked who the "leftovers" were, she noted that they were "all minorities, Middle Eastern, Asian, African American, Hispanic/Latino minorities." These examples show how alter‐provided guidance about what to expect meant that students were not blindsided when others treated them poorly. They were better prepared to deal with the uncomfortable situations they encountered.

Some students received guidance or opportunities that helped them feel included from organizations in which they participated. Students shared how they developed relationships with alters by participating in organizations, including preengineering organizations (i.e., STEM bridge programs), engineering‐related activities and design teams, and race/ethnicity‐ or gender‐focused organizations outside of engineering. The missions of such organizations typically include mentorship, guidance, and other professional development and relationship‐building opportunities. Women and URM students identified these opportunities as ones that helped them fit in their programs and persist in engineering.

Students reported that bridge programs (summer programs often held before the first year of college during which students take courses and engage in learning communities) helped them feel connected to a community of peers, a community they may not have otherwise encountered in their engineering program. For example, a Black woman enrolled in a PWI indicated that she initially did not feel like she belonged in her engineering program due to differences between her high school program and the types of high school programs to which others had access. She connected the value of the bridge program to surviving the crucial first years of the engineering program, which are particularly rough:

It's mainly the prerequisite courses, probably why so many younger people drop out ... [It] turn[s] students against each other, so no one wants to help each other. It's hard to find help and form study groups ... It was challenging ... it's hard to find a place to fit ... [It's] not like that in your major classes, ... I didn't know anybody when I came here, so my university's bridge program helped me find community before school started.

She noted that her participation in a bridge program helped her meet people, which provided her a sense of belonging, and hypothesized that students might leave the engineering program if they did not feel like they fit into the competitive environment. Numerous participants said having noncompetitive engineering peers with whom they could study, ask questions, and receive advice was a valuable resource for fitting in and persisting in their engineering program.

Students also negotiated their feelings of fit with greater ease when they cultivated relationships with alters through engineering‐related organization activities, such as design competitions. For instance, a White woman from a PWI said she "stuck out" more than she "fit in," but noted that an engineering‐related program helped her be more comfortable. She stated:

I fit into the aerospace department because of my university's aircraft design team. They're inclusive, supportive, and they taught me almost everything I know about airplanes. Being a part of a team, that made me feel included. I don't know where I would be if I didn't have a team like that ... Without hands‐on experience ... how are you going to be prepared in the real world? I fit in with my team.

In this specific case, the student not only increased her feelings of fit, but she also gained valuable engineering knowledge by participating in the design team.

Students said that cultivating relationships in clubs or organizations outside of engineering that were race/ethnicity‐, gender‐, and/or academic‐focused also helped with their fit in engineering. For example, a Latinx man from an HSI who joined a Latinx fraternity said:

There weren't a lot of engineers in there, but it helped me [to] have ... somebody to struggle with ... somebody to look up to ... I figured that if I had a lot on my plate, it would help me really get my stuff together ... really make me manage my time...

He explained, "I don't know if it's being homesick or missing talking Spanish or something; it feels a little off." Being around other Latinx students helped him combat his feelings of isolation.

Students who lacked relationships like those documented above reported feeling less fit in engineering. For example, a Black man from a PWI said he did not feel like he fit in his engineering program because he did not have any community in professional organizations through the college nor in student clubs for engineers or other organizations through resident life. He indicated that he felt that he simply "serve[d] as a statistic" for his program.

Previous research has demonstrated the link between fit and STEM major declaration and/or persistence (Fink et al., 2020; Meyer & Marx, 2016; Pugh et al., 2021; Rainey et al., 2018; Rockcliffe, 2020). Our first research question asked about how women and URM engineering undergraduates fit into their program. We found that participants generally had a good sense of whether they felt they fit or not. Black students were overrepresented among those who did not fit, with 31% of the interviewee sample being comprised of Black students, but Black students making up 63% of those who did not fit.

We found that women and Black students' feelings of isolation due to being a numerical minority affected their fit (Cech et al., 2011; Malone & Barabino, 2009). Women of all races/ethnicities and Black men discussed being one of the few people like themselves. We found that Black women experienced the double bind of both sexism and racism, increasing their net vulnerability, consistent with previous research highlighting their hypervisibility (Blosser, 2020; Dancy et al., 2020; McGee, 2020; Morton & Parsons, 2018; Rainey et al., 2018; Spencer, 2006).

At the same time, we found that women and URM students' experiences in their engineering program were mitigated by the extent to which they held statuses closely aligned to those of the dominant group based on their physical (i.e., eye color) and social characteristics (i.e., language repertoire). Specifically, White women and Latinx students tended to fit. The identities do not actively cause the observed phenomena (i.e., better fit), instead the cause is relational structures in service of actively marginalizing those identities. These statuses possibly represent a low proxy measurement of experiencing discrimination and lack of representation. However, such benefits to fit associated with being male did not extend to Black men, whom previous research have shown face tribulations in STEM (McGee, 2020). Also, in line with the exclusive STEM culture, many students who fit did not consider how their roles and behaviors or the behaviors of those from the dominant group impacted the experiences of women and Black students.

Regarding the experiences that influence fit (our second research question), we found that gender and racial stereotypes came from a range of people students encountered in their program, affecting students' feelings of fit. This finding is consistent with Leyva et al.'s (2021) results, which found connections between stereotyping fit in mathematics for women and URM undergraduates. It also aligns with Rockcliffe (2020), which linked Black women's experiences of racial microaggressions with a decreased sense of fit in engineering. We also found that women are confronted with anti‐women sentiments that cast doubt on their intelligence, a problem magnified for women considered to be beautiful. In addition to dealing with such prejudice, women had to deal with resource restrictions, whereby men prevented them from using power tools and, consistent with previous research, in engaging in hands‐on group work (Faulkner, 2009).

Like Rainey et al. (2018), we found that men did not recognize the influence of gender stereotypes on women. Consistent with previous research that showed that men were unaware of how race or gender affected how individuals are treated and fit in STEM (Dancy et al., 2020), men in our sample tended to explain and minimize disparities in the underrepresentation of women in STEM by adopting non‐structural explanations for these differences (i.e., that women are simply not interested in STEM). Women also reported that men often attempted to gaslight them by offering alternative interpretations of others' actions that did not align with women's perceptions (i.e., explaining away gender microaggressions).

Black students reported experiencing the poorest fit, worst treatment, and most hostile engineering program environment. This matches Riegle‐Crumb et al. (2019), who suggested that, after controlling for socioeconomic background, institution type, and high school academic preparation, Black students' increased likelihood of STEM departure in comparison to Latinx students' was likely affected by patterned stereotyping and microaggressions against Black students. Similar to Lee et al. (2020), Black students in our study reported the need to confront anti‐Black attitudes on campus all the time. These experiences influence their feelings of belonging in their engineering programs and beyond. Black participants' attitudes showed evidence of Black X Consciousness (Morton et al., 2019), in which there were collective practices in support of Black student identity and success. However, Black student responses but become more mixed when it comes to gender, testing the extent to which a shared Black X Consciousness might permeate patriarchal ideologies.

Like how gender stereotypes were not a specific concern for men, racial stereotypes were not a specific concern for White women or Latinx students (see also Dancy et al., 2020). We do not claim that Latinx students do not face structural and interpersonal obstacles and discrimination related to ethnicity. However, the different experiences reported by Latinx, US‐born Black, and non‐US‐born Black students do illustrate a racial hierarchy, in which Latinx and non‐US born Black students can experience elevated treatment in contrast to their American Black peers (Bonilla‐Silva, 2006; Saucier & Woods, 2016), evident of anti‐Blackness in STEM (Bullock, 2017; Cedillo, 2018; Martin et al., 2019; Nxumalo & Gitari, 2021; Vakil & Ayers, 2019).

A major contribution of our study is the focus of our third research question on the specific processes through which network‐based and participatory alters may affect fit. For most participants, alters, in general, helped their sense of fit improve with time. In accord with our hypothesis, the ways that network‐ and participatory‐based alters affected fit differed. We posit that this is likely due to differing closeness in social proximity between relationships with people in an individual's own network versus people they meet through participating in organizations. Specifically, we found that network‐based alters provided more personal warnings to students about the gender‐ and/or racially driven stereotyping they might encounter. They offered insights about people's negative perceptions about their academic ability or helped them make sense of their experiences that had occurred (i.e., reassuring women that perceiving sexism was not just "just in their head"). These warnings prepared women and Black students, allowing them to better deploy protective mechanisms against racism, sexism, and harassment in STEM to manage how others perceive them, manage how others treat them, and mitigate their feelings of fit. In addition, these alters provided information helping women and Black students articulate fit in STEM in light of how others react to their identities in STEM's exclusionary culture (Morton & Parsons, 2018). Having access to explicit information about the likelihood of negative experiences served as armor that allowed students to focus on their goals and "prove others wrong." These findings extend prior research (i.e., Blakley, 2016; McGee & Martin, 2011; Nguyen et al., 2021) by showing how network‐based alters mitigate fit. Our work also reinforces the notion that alters are a crucial source of strategies that students, particularly women and Black students, can use to manage stereotype threats.

For participatory social capital, women and URM students indicated that interacting with or being part of an affinity group positively influenced feelings of fit, which supported their persistence. Participatory alters helped them feel connected to a communityor family (Fernández et al., 2021), feel like they belong, and be around people culturally similar to them. Although being connected to alters with direct connections to the field of engineering is beneficial, we found that students also benefitted from guidance provided by supportive alters with no connections to their disciplinary field. This finding is consistent with those in Espinosa (2011) and Smith et al. (2021), which highlight the importance of participatory social capital in STEM organizations to feelings of fit for women and URM students. Our work echoes the assertion that fit via professional role confidence can be aided through "verification from relevant others in one's social milieu (Burke, 2004; Cooley, 1902; Huntington, 1957; Rosenthal et al., 2011)" as cited in Cech et al. (2011, p. 648).

We identify these understandings of network‐based and participatory social capital as especially crucial in the first year when most students who switch out of STEM do so. Our findings allow the social theorization informing how social structures and social location of alters intersect to affect persistence depending on the time deployed. The specific effects of homophilious alters (particularly those who are network‐based, at least initially) on stereotype management are that alters help students negotiate their racial/ethnic and gender identities with STEM's culture of exclusion. We add the findings articulated here, particularly those on how network‐based alters warned students to expect discrimination, to those of our previous work. In another study, we analyzed the same interview data with the same 55 women and URM students used in this study and found that parents encouraged men and women and students of all races/ethnicities to declare and persist in their engineering major by reminding them that they can succeed (Puccia et al., 2021). Yet, in our analysis of a larger data set of 2186 engineering undergraduates who completed surveys, of which the 55 interviews comprise a subsample, we found that Black students may lack initial social capital as evidenced by the smaller sizes of their social networks (Skvoretz et al., 2020). Our later analyses of that same survey data showed that, when available later in their academic career, Black students excelled at accessing participatory social capital that would help them be around homophilious alters through professional engineering organizations (Smith et al., 2021). Thus, while Black students, when they have access to network‐based social capital, benefit greatly from the advice provided by alters that mitigates their fit, they may have less access to this capital. Therefore, bridge programs and other programming received early on can sustain them and encourage participation in professional engineering organizations to promote their persistence in engineering (Smith et al., 2021).

In sum, we provide evidence that network‐based and participatory social capital uniquely contribute to students' fit and, consequently, persistence in engineering. The presence or absence of negative stereotypical experiences and alters to help students deal with them directly impacts students' perception of fit in engineering in ways that articulate with constellations of intersectional positionalities (Morton & Parsons, 2018). Our research shows that women and Black students, in particular, bear the emotional burden of managing stereotypes or exerting effort to convince others that they belong—effort and time they could be spending on their studies (Lee et al., 2020; McGee, 2020). Thus, it is not simply the patterns, due to interactional structures, in which women and Black students face a discrimination tax. It is also that other students, who are not facing this tax, experience a patterned, structural advantage in which they are more accepted. We noted that warnings about stereotyping from network‐based alters, and opportunities to cultivate feelings of belonging in participatory social capital activities, were powerful fit mitigators for students with such social capital. When conceptualizing social capital at the patterned level and considering it as a structure, we could see how patterns in differential treatment were linked to differences in social capital across time points to shape variations in widescale fit among groups influential in persistence.

Below we highlight three critical implications for practice. Additionally, we note that obstacles to full engagement in their engineering programs are not ones that women and URM students can address as individuals. Instead, they are systemic issues that institutions, and more specifically, engineering programs, must address. Keeping in mind women's increasing (but still markedly disparate) and Black students' somewhat stagnant representation in engineering (Roy, 2019), the context described by participants illustrates the difficult terrain that they must navigate. This context also suggests that increasing the rate by which women and URM students declare engineering as a major alone is inadequate to improve their persistence. First, given the documented effect of social capital on fit, we note the critical need for universities to implement structural programming early during their first year to improve women and URM students' access to social capital, particularly for students with lower entry social capital. For example, research has shown that women and URM students are often unaware of "unofficial" routes to STEM success (McGee, 2020; Seymour, 1999; Stevens et al., 2008). Therefore, activities and interventions that provide women and URM students with additional social connections provide critical resources and knowledge and can contribute to their persistence (Atman et al., 2008; Daily et al., 2007; Goodman et al., 2002; Shapiro & Sax, 2011; Stevens et al., 2008). Importantly, universities must use nuanced approaches when considering URM students' experiences as a collective group. Specifically, universities must provide Black students with additional support given their differential experiences compared to those of other URM groups (Lee et al., 2020; McGee, 2020) and because a lack of perceived fit can impact ability to obtain social capital (Blosser, 2020).

Second, because individuals with majority statuses may not acknowledge gender and racial prejudice and discrimination, care needs to be taken in interventions seeking to build student capital. For example, it will be essential to connect students to alters who have similar statuses (Kricorian et al., 2020). Blosser (2020) notes that mentorship experiences for URM students can be facilitated by creating a critical mass of minoritized students so that students are able to develop relationships with others like themselves, including in counterspaces that are safe for minoritized groups (Ong et al., 2018), or by faculty playing a larger role in creating student study groups. However, due to engineering's lack of diversity, institutions often experience difficulty when attempting to provide homophilious mentorship for women and particularly URM students (McGee, 2020). If such experiences are not possible in local institutions, McGee (2020) asserts that alters demographically dissimilar from students can still provide relevant advice if they recognize how others' gender and racial/ethnic stereotyping affects students. The design of participatory‐based social capital opportunities, such as earlier engagement in professional STEM society conferences that are inclusive, as well as buttressing diversity, equity, and inclusion efforts in such societies (Peters et al., 2021), are suggested approaches to help students to connect with possible mentors who are like themselves (Campbell‐Montalvo et al., In press; Campbell‐Montalvo et al., 2020; Smith et al., 2021).

Third, universities often attempt to address STEM inequality through individualized, deficit‐oriented approaches and locate problems within students (Peck, 2020), leaving intact racist structures (McGee, 2020; McGee et al., 2021). To combat this, we encourage universities to design interventions for members of dominant groups, both faculty and students, that problematize, recognize, and raise awareness of stereotyping, microaggressions, and other ills (e.g., implicit bias). Interventions should support majority groups recognition of the influence of their words, behaviors, and actions on minoritized groups and their feelings of fit (Blosser, 2020). Because raising awareness does not directly change behavior, policy and structural changes in the university must accompany efforts to improve how members of the dominant group treat minoritized students. Taking such an approach challenges the narrative that positions members of minoritized groups as problems to be addressed. Specifically, interventions should encourage members of the dominant group to hold themselves and others accountable (McGee, 2020), and teach them how to be effective allies and intervene and be supportive when appropriate (Moore & Nash, 2021). We add that respondents also shared examples of prejudice and discrimination experiences in K‐12 settings and the workplace during internships, suggesting that interventions are likely also needed before and after students enroll in college. See also Fletcher et al. (2017) for additional information on how this issue can be addressed by professional organizations, industry, policy, K‐higher ed, family, religious organizations, and research.

We identify two limitations to the interpretation of our results. First, we recruited participants from engineering programs when they were enrolled in the spring semester of their first year or the fall semester of their second year. Thus, our sample may overrepresent individuals who were more likely to persist in engineering, perhaps obscuring perspectives of students struggling with fit and preventing us from seeing the role social capital may have played or not played in their experiences. Second, our interview team's composition—two Black women and one White woman—may have influenced the nature of students' responses. Yet, our results show that participants discussed racial and gender stereotypes in great detail.

As noted by a participant and established in previous research, the hypercompetitive and "weed out" culture may encourage students to consider alternatives to engineering (Seymour & Hewitt, 1997; Weston, 2019). Therefore, students' perspectives during their early years of matriculation (i.e., their first semester) must be captured to provide understandings about their experiences at such a critical time. To build on earlier work with brief interventions (Jordt et al., 2017; Walton & Cohen, 2011; Yeager et al., 2016), we encourage the mixed‐method study of more robust, sustained interventions, including those aimed at women and URM students as well as at dominant groups. Researchers must consider such interventions from multiple perspectives to capture the complexities and differential influence that may result across time. Insights can support broadened engineering and broader STEM participation and persistence.