This chapter discusses the topic of identity and how ignoring a person’s identities and sociodemographic background, including first-generation (FG) status,1 without positively recognizing and affirming the value of differences, can affect specific populations of mentees in White, male-dominated science, technology, engineering, mathematics, and medicine (STEMM) disciplines. Particular attention is given to underrepresented (UR) students and FG students, as well as sexual- and gender-minority students, and students with both visible and nonvisible disabilities.2 For students with these identities, access to social capital, cultural capital, and networks through both family background and mentorship is oftentimes more limited than that of their peers (Pascarella et al., 2004).3 In addi-
1Sociodemographic refers to social and demographic factors such as race, ethnicity, age, sex, gender, sexual orientation, socioeconomic status, (dis)ability status, religion, education, migration background, and culture. First-generation students are the first members of their families to attend college.
2 This report refers to UR groups as including women of all racial/ethnic groups and individuals specifically identifying as Black, Latinx, and American Indian/Alaska Native. Where possible, the report specifies if the UR groups to which the text refers are Black, Latinx, or of American Indian/Alaska Native heritage. Sexual- and gender-minority students refers to students with identities that include sexual orientation identities such as lesbian, gay, bisexual, queer, and asexual, as well as gender identities such as pre- and posttransition transgender, intersex, and nonbinary. Students with nonvisible disabilities include students with identities such as autism spectrum disorder, attention deficit hyperactivity disorder, dyslexia, and other neurodiverse conditions.
This chapter provides an overview of factors that can affect aspects of different identities, as well as the role that mentorship can play in building an identity that is connected strongly to science—a science identity—and that does not undermine other distinct visible and invisible attributes of identity, such as culture, race, gender, and ability status. It also reviews the evidence supporting the idea that mentorship of UR students can play a critical role in addressing their underrepresentation in STEMM. While many of the topics discussed in this chapter are relevant to multiple identities—and may be presented in generalized terms—the committee stresses that the discussions here should not be understood as disregarding the intricacies of any particular identity or the differences between identities.5 Instead, the intent for this chapter is to raise awareness and motivate mentors to engage in introspection and do “self-work” as a means of becoming more effective in their mentoring relationships with their diverse mentees.6 Additionally, this chapter provides UR mentees with a vision of how to see themselves in the context of STEMM and potentially recognize some of their own experiences.7
THE IMPORTANCE OF IDENTITIES
Faculty working with undergraduate and graduate STEMM students in classrooms and research environments are interested in sharing knowledge, providing training,
4 In recognition of intersecting identities, intersectionality is the term that is used to acknowledge, account for, and conceptualize “multiple grounds of identity” (Crenshaw, 1991). It is the complex, cumulative way in which the effects of multiple elements of identity (such as race, gender, and class) combine, overlap, or intersect, especially in the experiences of marginalized individuals or groups.
5 Where possible, details about specific studies are provided.
6 While this chapter provides the reader with an introduction to understanding identities, Chapter 4 provides examples of structures and approaches to mentorship, and Chapter 5 provides educational resources that can be utilized to appreciate different identities as a means of continuously improving one’s mentoring practice.
7 A representative, but not exhaustive, list of programs that include mentoring experiences, some of which focus on supporting UR mentees in their pursuit of academic and career goals, is included in Appendix B.
accelerating discovery, and facilitating students’ preparation for STEMM careers. Increasingly, as universities expand their missions to better recruit and retain students from diverse backgrounds, faculty have questions about how to best engage in mentoring relationships with students who come from backgrounds different from their own (Clayton-Pedersen et al., 2017; HHMI, 2016).
The National Academies report Expanding Underrepresented Minority Participation: America’s Science and Technology Talent at the Crossroads (NAS-NAE-IOM, 2011a) made the case for why increasing the number of individuals from groups currently underrepresented in the STEMM workforce is vital to the nation’s interests, namely, to expand economic opportunity to all members of the nation’s population and to meet the growing demand for STEMM-trained professionals (U.S. DOC, 2017). A more recent National Academies report, Minority Serving Institutions: America’s Underutilized Resource for Strengthening the STEM Workforce (NASEM, 2019), reiterated this message and noted that increaasing workplace diversity grows the available talent pool and brings a broader range of perspectives and expertise to bear on solving grand challenges in STEMM. STEMM workplace diversity also boosts work performance and engagement, improves research quality and health care, and fosters innovation and growth (Cohen et al., 2002; Federal Glass Ceiling Commission, 1995; Florida, 2014).8
As noted in Chapter 1, one of the best ways to develop the STEMM workforce is to educate and train the full diversity of students (PCAST, 2012). Mentoring students from diverse backgrounds can help cultivate STEMM professionals with different perspectives who will assist with scientific competition, collaboration, enhanced creativity and problem solving, learning, and effectiveness (Bert, 2018; Mannix and Neale, 2005; NIH, 2019; Summers, 2011, 2012).
Some progress has occurred since the Expanding Underrepresented Minority Participation report was published. However, as of 2017, women, persons with disabilities, and members of three racial and ethnic groups—African Americans, Latinx, and American Indians or Alaska Natives—as well as FG college students are still underrepresented in educational attainment and the STEMM workforce (Espinosa et al., 2019; NASEM, 2018a, 2018b, 2018c, 2019; NCSES, 2017; U.S. DOC, 2017; U.S. GAO, 2017).
Despite widespread recognition that a lack of diversity among STEMM practitioners deprives the nation of involving all segments of the population in what are projected to be among the fastest-growing sectors of the economy, a variety of factors keep undergraduate students from UR groups from choosing and remaining in STEMM disciplines. At the graduate level, underrepresentation is even more pronounced (NASEM, 2018c; Weddle-West and Fleming, 2010). Although there have been improvements in diversifying STEMM programs, many scholars point to effects of race and racism in STEMM, which lead UR students to feeling alienated, having to work twice as hard to receive recognition, and working under constant scrutiny (McGee, E. O., 2016; McGee et al.,
2019). These feelings may also result from implicit biases of mentors or fellow students,9 in which attitudes or stereotypes about UR students affect how they are treated even in the absence of explicit racism (Burt et al., 2018).
Although FG college students account for one-third of all students entering postsecondary education—and almost half of all students enrolled at minority-serving institutions (Harmon, 2012)—they are less likely than continuing-generation (CG) students to begin their studies in 4-year colleges and more likely than CG peers to attend less selective colleges, including 2-year and for-profit institutions (Cataldi et al., 2018).10 They are less likely to have taken a college preparation curriculum, and only 20 percent of FG college students obtained a 4-year degree 10 years after their sophomore year of high school compared with 42 percent of CG students (Redford and Hoyer, 2017). FG students also leave science, technology, engineering, and mathematics (STEM) majors at higher rates than CG students (Shaw and Barbuti, 2010).
Additionally, FG college students tend to come from the lowest income quintiles (77 percent, of which 27 percent come from a household income of $20,000 and under and 50 percent come from a household income of $50,000 and under) and are more likely to be Black (11 percent) or Hispanic (27 percent) than CG peers (Redford and Hoyer, 2017). They face particular and unique challenges that often intersect with the identity-based challenges regarding academic preparation that come with a background created through political and historical processes to have a particularly devalued status and the expectation of assimilation into the dominant culture of higher education. Lower levels of family financial support along with different expectations and career goals contribute to FG students being more likely to drop out after or during the first year, significantly less likely to complete an undergraduate degree in 6 years, and less likely to enroll in graduate programs than CG students (Richardson and Fisk Skinner, 2006; Warburton et al., 2001). These students may not have the same advantages that come from the cultural and financial capital of college-educated parents to help navigate college, posing unique challenges in preparing them for STEMM careers and integrating them in research-based mentoring models that assume knowledge about careers and academic success.
Addressing the underrepresentation of major segments of the nation’s population will require a multipronged approach, but mentorship will likely constitute a significant component of the complex solutions that are required. Numerous studies have shown that effective mentorship for UR students enhances recruitment into and retention in research-related career pathways (Bhatia and Amati, 2010; Dasgupta and Stout, 2014; Dennehy and Dasgupta, 2017; Hathaway et al., 2002; Nagda et al., 1998; Ong et al., 2011). Research on undergraduate students shows that mentors play a critical role in
9Implicit biases are “attitudes or stereotypes that affect [the holder’s] understanding, actions, and decisions in an unconscious manner. These biases, which encompass both favorable and unfavorable assessments, are activated involuntarily and without an individual’s [conscious] awareness or intentional control” (OSU, 2015).
10Continuing-generation students are students that have at least one college-educated parent.
Despite the positive effect that mentorship has on UR students, studies have reported that UR individuals enrolled in STEMM degree programs typically receive less mentorship than their well-represented peers (Gayles and Ampaw, 2011; Helm et al., 2000; King et al., 2018; Thomas, 2001; Thomas and Hollenshead, 2001). Indeed, research shows that UR students’ mentorship requests for mentoring meetings are more often ignored by mentors than those of White men (Milkman et al., 2015). Regarding FG college students, White FG students place more limited value on having a personal connection with one’s mentor than African American FG students (Ishiyama, 2007). White FG college students also view personal and career development as key mentoring benefits, while African American FG students saw career clarification as the most crucial mentoring benefit.
WHAT IS IDENTITY?
Identity is the composite of who a person is. Identity includes the way one thinks about oneself, the way one is viewed by the world, and the characteristics that one uses to define oneself, such as an individual’s gender identification, sexual orientation, place of birth, race, ethnicity, FG college status, profession, values, and even hobbies (Crenshaw, 1991; Felix-Ortiz et al., 1994; Hall, 2014; Hall and Burns, 2009; Helms, 1990; Jones and McEwen, 2000; Nash, 2008; Sellers et al., 1998; Shields, 2008). Some aspects of identity are constant, while others change depending on stage of life and social context. In addition, a person can hold multiple identities that also intersect one another, such as Black, transgender woman, scientist, spouse, parent, artist, bookworm, and athlete. Research on the persistence of UR populations has often highlighted specific aspects of identity such as race, ethnicity, gender, income, and FG status as particularly important factors in retention and success in college and in STEMM fields (Archer et al., 2010; Calabrese Barton et al., 2013; Kim et al., 2018; Merolla and Serpe, 2013; Stephens et al., 2014). Identity can also govern access to social capital and network resources, and affect power in relationships.11
DEVELOPMENT OF IDENTITY
Individuals develop social identities to fill psychological needs, such as increasing self-esteem (Reid and Hogg, 2005) and reducing uncertainty about oneself (Hogg and Mullin, 1999). Developing social identities requires both a sense of belonging to a particular social group and recognition as an accepted member of the group from existing members of that social group. Accordingly, social identities are defined by a common
set of norms, attitudes, traits, and stereotypes that together form a “prototype,” the typical or average representation of a group member (Hogg et al., 1995). Individuals who deviate from this prototype—in STEMM, those individuals who are not White, male, heterosexual, able-bodied, middle-class and up, or otherwise historically represented as scientists—are more likely to be marginalized within the social group and not extended full membership. This marginalization, sometimes in the form of microaggressions,12 has the effect of barring UR students from benefiting fully from opportunities afforded to members of more well-represented and prototypical groups. Student experiences in STEMM contexts are highly contingent upon their social identities (Kim et al., 2018; Tajfel, 2010; Tajfel and Turner, 1986), so marginalization in and ostracization from STEMM social groups can challenge the process through which emerging scientists who may not “look the part” develop a social identity as a scientist (Kim et al., 2018).
The concept of science identity includes social and cultural identity constructs to explain how an individual can develop a professional identity in the culture of science (Byars-Winston and Rogers, 2019).13 An individual assumes and nurtures a science identity by developing scientific competence in one’s own mind and in the eyes of others, by having the skills and opportunities to act like a scientist, and by obtaining recognition from oneself and meaningful others as being a scientist (Carlone and Johnson, 2007).14 Being recognized as a scientist by meaningful others is a critical component for developing a science identity (Carlone and Johnson, 2007). Science identity can also be reinforced by cultural communities that internally acknowledge a scientist in that role (Chemers et al., 2011; Rodriguez et al., 2019).
UR students’ mentored research experiences strongly correlate with their sense of science identity, particularly for African American men (Byars-Winston and Rogers, 2019). The unique gendered-racialized experiences of African American male students in STEMM, such as encountering gender-specific racial stereotypes, isolating institutional practices, discrimination from non-Black peers, and race-based faculty biases, can lead to
12Microaggressions refer to “the everyday verbal, nonverbal, and environmental slights, snubs, or insults, whether intentional or unintentional, which communicate hostile, derogatory, or negative messages to target persons based solely upon their marginalized group membership. In many cases, these hidden messages may invalidate the group identity or experiential reality of target persons, demean them on a personal or group level, communicate they are lesser human beings, suggest they do not belong with the majority group, threaten and intimidate, or relegate them to inferior status and treatment” (Sue, 2010).
13Science identity refers to a professional identity within the scientific culture. A cultural identity is a social identity that is associated with a nationality, ethnicity, religion, social class, generation, or any group defined by a distinct culture. These connections of identities to STEMM professions have origins as a conceptual model for the career development of women of color (Carlone and Johnson, 2007). This conceptual model describes how recognition of one’s self and others’ recognition of them as a potential scientist becomes their career-related identity (Pfund et al., 2016). Career-related identity is an important factor in predicting some future science-related behaviors (Carlone and Johnson, 2007; Seymour et al., 2004; Vincent-Ruz and Schunn, 2018; Williams and George, 2014).
14Meaningful others refers to people an individual identifies as those from whom acceptance matters.
role strain and self-doubt due to negative advising experiences for this population (Burt et al., 2018). One study found that mentors taking time to help African American male students in STEM work through their research tasks had a statistically large influence on the students’ science identity, research self-efficacy, and research career intentions (Bidwell, 2015; Byars-Winston and Rogers, 2019). These studies underscore that faculty acknowledging social identity within a nurturing relationship is important for male African American students.
In addition, science identity is linked strongly in many contexts to a sense of self-efficacy (Hunter et al., 2007; Steiner et al., 2004), an individual’s belief in their capacity to attain specific performance goals in science (Bandura, 1997; Byars-Winston and Rogers, 2019).15 However, longitudinal studies of UR undergraduate students have shown that self-efficacy alone does not predict persistence and integration into the scientific community. Rather, persistence and integration appear to require an individual to develop a science identity and internalize scientific values (Estrada et al., 2011).16 Science identity, however, does not predict advancement into medical school (Cruess et al., 2014; Frost and Regehr, 2013; Goldie, 2012; Wilson et al., 2013).
Research has shown that how an individual’s science identity fits with other social identities, such as gender, race, or socioeconomic status, has a significant effect on career goals (Carlone and Johnson, 2007; Chang et al., 2011; Chemers et al., 2011; Estrada et al., 2011; Hurtado et al., 2009). These studies show that graduate students use a variety of strategies to develop and integrate their science identity and other social identities. For example, they may define their own sense of what it means to be a scientist and a “person of color.” They might also create ways of simplifying science to make it more relevant and accessible to their nonscientist friends and family members or manage different identities in different contexts (Tran, 2011). Moreover, some “hidden” identities, such as sexual and gender orientation, socioeconomic class, and FG status, make certain issues more difficult to address in STEMM because students do not always reveal that these identities require support.
TENSIONS WITH IDENTITY IN STEMM CONTEXTS
Many UR and FG students experience STEMM contexts differently than their peers whose identities are well represented, either because of persistent social and racial stereotypes (McGee, E.O., 2016) or as a result of unclear communication from faculty regarding strategies for student success (Burt et al., 2018; Davidson and Foster-Johnson, 2001). For example, one study found that negative racial experiences in the first year of college tend to negatively affect the otherwise positive relationship between developing a science identity and persisting in STEM (Chang et al., 2011). UR students may also
experience stereotype threat, or the risk of conforming to common, negative stereotypes about gender or race, that can negatively affect their academic performance.17 Often, UR scientists must balance more social and cultural identities that differ from the prototype of a person in STEMM compared with well-represented scientists (Brown, 2004; Carlone and Johnson, 2007; Johnson et al., 2011; Ong, 2005). Similarly, FG students may struggle to reconcile their family and home identities that are socially and culturally distinct from the college environment with those required for success in college (Orbe, 2008).
More broadly, UR students’ awareness of how society and schools position them as underachieving influences how they construct their career-related identities (McClain, 2014). One study with UR STEMM doctoral students found that although many of these students could perform scientific research competently, they lacked recognition from peers and supervisors as legitimate and competent members of their scientific communities, resulting in alienation from the laboratory community and even dampening of their STEMM career consideration (Malone and Barabino, 2009). Preliminary findings provided by Vanderbilt University’s basic biomedical sciences Ph.D. program show that over nearly 20 years, Hispanic and White students received comparable student performance evaluations from their mentors, while African American students were evaluated on average only 50 percent as positively (Brown et al., 2019).18 Because individuals cannot construct a social identity in the absence of recognition from others, feeling invisible can thwart the development and reinforcement of a person’s science identity. Women from UR racial or ethnic groups with “disrupted identities,” for example, have reported that their bids for recognition, and thus the development of their science identities, were unsettled by others’ interactions with them.19 These interactions were shaped largely by those individuals’ perceptions of who does and does not belong in science based on race, ethnicity, and gender (Carlone and Johnson, 2007).
What Is Identity Interference?
Research has shown that UR students are often expected to conform and assimilate into the dominant White, male culture and minimize their race- and gender-informed identities (Davidson and Foster-Johnson, 2001). Although it is ideal to unify one’s various identities, particularly for emerging adults (Erickson et al., 2009; Erikson, 1968), com-
17 While scholarship on stereotype threat has shown its impacts via academic, psychological, and even interpersonal measures, the effect of mentorship on reducing stereotype threat has not yet been studied (Cromley et al., 2013; Holleran et al., 2011; Steele and Aronson, 1995; Thomas and Erdei, 2018). Emerging evidence suggests that mentoring practices aimed at reducing stereotype threat may contribute to fuller representation of individuals from underrepresented groups in the sciences (Byars-Winston et al., 2018; Estrada et al., 2017).
19 Women with disrupted identities had career trajectories that were “rockier, most unstable, and less satisfying,” a result of lack of recognition from meaningful others (Carlone and Johnson, 2007, p. 1197).
partmentalizing one’s identities is often the case for UR students in STEMM, reflecting an underlying process called identity interference (Settles, 2004). Identity interference occurs when cultural meanings and stereotypes assigned to social identities cause those with multiple identities to feel that one identity interferes with the successful performance of another identity.20
For UR students, identity interference means they often maintain separate social and academic peer networks (Malone and Barabino, 2009; Tate and Linn, 2005), minimize displaying their race- and gender-informed identities, and compartmentalize rather than integrate these critical identities with their science identities (McCoy et al., 2015). The same is true for those with minority sexual or gender identities (Flanagan, 2017; Puckett et al., 2016; Yoder and Mattheis, 2016). Resolving this interference by disidentifying, minimizing, or downplaying their devalued social identity can in turn challenge students’ sense of authenticity and sense of belonging in their discipline (McGee, E. O., 2016; Roberts et al., 2008; Settles, 2004) and can even result in lower academic or professional performance (Darling et al., 2008).21 Students who feel they must change themselves and their identities to fit in are more likely to experience depression, reduced psychological well-being, and impaired academic performance (Roberts et al., 2008; Settles, 2004). Extensive empirical evidence confirms the tensions that can arise from being the “other,” the “only one,” or the “unknown” (Espín, 1991, 1997; Johnson et al., 2011; Malone and Barabino, 2009; Ong, 2005; Ong et al., 2011), simultaneously bringing invisibility to one’s identity as a scientist and too much visibility to one’s UR identity.
How Does Mentorship Help Develop Science Identity?
By contributing to the socialization and integration of students into scholarship and academe as a community, effective mentorship plays a critical role in developing a science identity (Byars-Winston et al., 2015; Eagan et al., 2011; Eby and Dolan, 2015; Estrada et al., 2018; Freeman, 1999; Gandara and Maxwell-Jolly, 1999; Gasiewski et al., 2012; McGee and Keller, 2007; Robnett et al., 2018; Thiry and Laursen, 2011), which then makes it more likely they will continue on in STEM fields after graduation (Barlow and Villarejo, 2004; Estrada et al., 2011). Mentorship also helps students see themselves as STEMM scholars who can contribute to their disciplines (Wilson et al., 2012). Given that developing a science identity is a strong and unique predictor of who will continue on to graduate school in a STEMM field, colleges and universities should enable experiences that help undergraduates feel they belong in and are included in the scientific
20 Organizational context can also affect the performance and perceived acceptance of identities. For example, studies have shown that minority-serving institutions often intentionally cultivate campus climates of belonging for students as a strategy for success (NASEM, 2019).
21 The effects of deemphasizing a devalued identity in terms of psychological and academic outcomes is worse for students whose racial identity is more central to their sense of self (Oyserman et al., 2012; Roberts et al., 2008; Settles, 2004).
culture, which in turn will help foster the development of a strong science identity and increase retention rates in STEM for UR undergraduate students (Estrada et al., 2018). Research has shown, for example, that the development of a science identity is predictive of an individual staying on a STEM career pathway for up to 4 years after graduation (Estrada et al., 2018).
For graduate student mentees, the psychosocial support functions of mentorship have been found to influence science identity (Chemers et al., 2011). Given that self-efficacy and science identity need to mesh with other aspects of social identity (Bakken et al., 2010; Hunter et al., 2007; Ries et al., 2009),22 mentors need to understand how various identities interact with one another in their mentees. Mentors should also accept that the identities of their mentees will likely evolve as they progress toward becoming STEMM professionals and continually assess their competence as STEMM professionals.
MENTORSHIP FOR UNDERREPRESENTED STUDENTS IN STEMM
Mentorship for UR students is vitally important to their success, but they are less likely than well-represented students to receive mentoring (Felder, 2010; Gayles and Ampaw, 2011; Johnson, 2015; King et al., 2018; Thomas, 2001). At the same time, approaches that help the general student body may not necessarily work best for UR students. While there have been improvements in diversifying STEMM training and education programs, many scholars point to the continuing effects of race and racism in STEMM, including reports of students feeling alienated, having to work twice as hard to receive recognition, and working under constant scrutiny and suspicions of presumed incompetence (McGee, E. O., 2016; McGee et al., 2019). These are also examples of microaggressions, implicit biases, and manifestations of stereotype threat.
Ignoring or being silent on these realities will not mitigate their effect. Effective mentorship requires that faculty have an awareness of the identity-related challenges their mentees may have, as well as a set of learnable skills, to effectively support the talent development of UR students in the context of racial realities in STEMM. In one study conducted with a sample of research mentors largely from well-represented backgrounds and undergraduate mentees largely from UR groups in STEMM, mentees were more likely than mentors to endorse having cultural diversity matters directly addressed in the mentoring relationship (Byars-Winston et al., forthcoming). This finding is ripe for further inquiry into the effect of mentors’ cultural awareness in research mentoring relationships and has implications for mentorship education to support mentors’ cultural responsiveness in their mentoring practices.23 In this section, we review research
22 Including race, ethnicity, sexual orientation, socioeconomic class, and gender.
supporting the positive influence on student outcomes from faculty engagement and intentionality in developing and enacting culturally responsive mentoring methods.24
How Does Identity Affect Mentorship?
Based on numbers, mentors in STEMM fields are typically White or Asian, and research shows that majority mentors are more likely to hold “colorblind” views of their students and to dismiss the idea that social identities shape their students’ academic experiences (Brunsma et al., 2017; McCoy et al., 2015; Melton et al., 2005; Prunuske et al., 2013).25 Some STEMM faculty from well-represented groups may espouse this ideology because of concerns of being misunderstood by or offensive to their mentees, not knowing what to say, or even fear of being perceived as prejudiced (Byars-Winston et al., 2019). Some UR faculty in STEMM, especially those at predominantly White institutions, may be likewise uninclined to directly address social identities and cultural diversity matters because of their own experiences with inequities in institutional roles and research support, being the one to whom more UR students turn for support, and fatigue from being overextended in service and teaching (Armstrong and Jovanovic, 2017; Xu, 2008).26 However, based on the evidence, ignoring race, gender, and other important social identities is to deny the formative effect of these identities on students’ experiences in their programs and later careers.27 For example, UR students may be less likely to ask questions if they do not feel they belong in a given environment in the first place.
Mentorship has the potential to ameliorate many identity-related challenges for STEMM students in higher education and perhaps even inoculate them against those challenges. Mentors from all backgrounds and in all contexts can work to acknowledge identities of their mentees and understand the research describing the influence of social identities on students’ experiences in STEMM. Studies have found that mentors who were
24Intentionality refers to a calculated and coordinated method of engagement to effectively meet the needs of a designated person or population within a given context.
25 Colorblind views include focusing exclusively on individual performance measures without consideration of factors that are highly correlated with performance such as social identities, cultural background, and social context. This tends to privilege individuals with better preparation, higher social capital, and fewer additional obligations—often White, male, single, full-time, non-FG students from higher socioeconomic backgrounds.
27 The appropriate level of focus on specific aspects of identity is dependent on individuals involved in the relationship and should be determined by the individuals involved during the establishment of the relationship (e.g., the initiation stage). Mentorship tools to assist with discussion during the initiation stage are discussed in Chapter 5.
culturally responsive—who had attitudes, behaviors, and practices that enable them to work with mentees with different cultural backgrounds (Sanchez et al., 2014)—and who understood power dynamics and oppression had success in fulfilling the needs of UR students (Felder and Barker, 2013; O’Meara et al., 2013). Culturally responsive mentorship can validate students’ various identities and help them navigate invalidating experiences they encounter while simultaneously reinforcing their self-efficacy in their field (Byars-Winston et al., 2015). This can greatly increase the likelihood of their thriving in STEMM environments (Thomas et al., 2007; Vaccaro and Camba-Kelsay, 2018).
Culturally responsive mentorship, whereby mentors show curiosity and concern for students’ cultural backgrounds and their non-STEMM social identities, may be one way mentors can validate their students’ multiple identities. In one study of White mentors who successfully engaged in cross-racial mentoring relationships with Black students at a predominantly White institution, the mentors reported (1) having a heightened awareness of the unique challenges facing Black students, (2) gaining a holistic understanding of the student, and (3) engaging in reciprocal relationship building (Reddick and Pritchett, 2015; Syed et al., 2011). While especially important in cross-racial relationships in White-dominated contexts, culturally responsive practices can also benefit mentorship when mentors are from marginalized communities themselves or at minority-serving institutions. Culturally responsive mentorship can also engage elements of identity beyond race. For example, deaf mentees rated their mentoring favorably if they perceived that their mentor was responsive to their deaf status, even if their mentor was not deaf (Braun et al., 2017).
Scholars who work on diversifying STEMM assert that faculty can improve their mentoring methods by considering contexts, styles, and the lived experiences of their students—including their social and science identities—in their actions to support students’ values and goals (Cropps and Esters, 2018; García and Henderson, 2014; Mondisa, 2015; Patton, 2009; Rasheem et al., 2018; San Miguel and Kim, 2015). A study of UR undergraduate STEMM students revealed that those reporting they had received culturally responsive mentoring also felt more confident as a researcher, refined their academic and career goals, and became more committed to graduate school and a graduate degree (Haeger and Fresquez, 2016).28 Together, these studies indicate that culturally responsive mentoring correlates with students feeling more attached to their field of study and to the research world.
Many interventions are designed to target multiple social identities as a group, such as low-income, UR, and FG students. Many studies also focus on the overlap of these particular identities. For example, compared with other racial/ethnic groups, Latinx
28 In that study, culturally responsive mentoring strategies included understanding how students’ backgrounds (e.g., ethnicity, gender, social class) contribute to their student experience; spending time getting to know them, their background, and their goals early in the research experience; and closely relating to their personal background.
college students are more likely to be represented among FG college students that also come from low-income backgrounds (Hurtado et al., 2007). Moreover, minority-serving institutions tend to attract and enroll higher numbers of low-income, FG, and UR students, and thus the success of those students is equivalent to the overall success of those institutions in producing STEM graduates (NASEM, 2019). Mentorship, including sponsorship, has been shown to be a strategy to promote student success at minority-serving institutions (NASEM, 2019).
Community-based peer mentoring among groups with shared identities can also play a role in affirming students’ identities and providing mentorship because of their multiple identities, not in spite of them. For example, two affinity-based professional societies—the National Society of Black Engineers and the Society for the Advancement of Chicanos/Hispanics and Native Americans in Science—work to affect change in STEMM underrepresentation. Students who attend the societies’ conferences begin their experiences with affirmation of their cultural identity through visual images that connect their cultural heritages to STEMM. They also benefit from preconference coaching that tells them they will be in an environment that will allow any participant to mentor another. In addition, there is an understood element of mentorship that occurs, and is expected, across peer groups—from high school to undergraduate to graduate to faculty and nonacademic STEMM professionals (Daily et al., 2007; Horwedel, 2005; Johnson and Sheppard, 2004; Litzler and Samuelson, 2013; May and Chubin, 2003).
How Does Race Affect Mentorship?
One study of mentoring experiences between White faculty mentors and Black college students found that White faculty saw mentorship as a “purposeful and iterative process of developing relationships with students” (Reddick and Pritchett, 2015). It may be that discussing or asking about students’ various non-STEMM identities could signal that the mentor recognizes and accepts various identities, or simply takes an interest in their background. For example, a study of cross-racial mentor-mentee relationships in a nonacademic work organization found that some Black mentees had highly salient racial identities and wanted to integrate their racial and professional identities and to openly discuss race with their mentors. When they were paired with a White mentor who held a colorblind perspective and preferred to suppress discussions of race and diversity,29 the mentees described receiving career support but not psychosocial support (Thomas, 1993). The mentees felt uncomfortable, said it was a barrier to forming a closer relationship, and reported they could not trust their mentor to make decisions based on race in a racially diverse workplace. Other investigators have confirmed this idea about White mentor–Black mentee dyadic pairings, finding that trust is more likely to develop when
29 Because of power dynamics, the mentors dictated this aspect of their relationship.
mentor and mentee agree on the significance or insignificance of race in the relationship and workplace (Blake-Beard et al., 2011).
One study found that unspoken assumptions about race and ethnicity can create problems even for those Black doctoral students and White faculty members who shared values of inclusivity (Gasman et al., 2004). The authors of this study concluded that faculty who work alongside UR graduate students should acknowledge that unequal power relationships and cultural forms of discrimination and oppression are common in academia. Institutions that fail to have faculty of any race or affinity-/identity-based student groups with whom UR students can discuss their interests create a strenuous and challenging experience for the students (Felder and Barker, 2013). Moving beyond racial boundaries requires mentors to leave their comfort zones if they want to build relationships based on honesty, equity, reciprocity, respect, and integrity (Gasman et al., 2004).
How Does Gender Affect Mentorship?
Research on women in cross-gender and same-gender workplace mentoring relationships suggests there may be more important factors that predict mentorship outcomes than gender similarity (Allen et al., 2005). For example, interpersonal comfort fully mediated the relationship between gender similarity in mentoring relationships and the mentees’ reports of the career and psychosocial mentoring they received. That is, although gender-matched pairs were more likely to report positive mentorship experiences, the correlation between gender matching and the positive experiences became insignificant when researchers measured how comfortable the mentees were interacting with the mentors. It appears possible, then, that finding ways to increase comfort across diversified mentoring relationships can improve the quality of those relationships. One study, however, found that female mentees reported more relational challenges with male mentors than female mentors, and male mentors reported more relational challenges with female mentees (Ensher and Murphy, 2011). No similar pattern was observed for male mentees and female mentors.
The perception of career support and psychosocial support that mentees received may also depend on the gender of the mentor. Early research on mentorship showed that female mentees with male mentors had difficulty seeing their mentors as suitable role models (Kram, 1985a) and that women in same-gender mentoring relationships reported significantly greater role modeling from their mentors (Ragins and McFarlin, 1990). Subsequent research has shown that both male and female students perceive female mentors as offering more psychosocial support, including role modeling, and male mentors as offering more career support, which is consistent with typical gender roles (Sosik and Godshalk, 2000). Similarly, research has shown that female mentees see male mentors as more focused on academic or career goals and female mentors as more focused on psychosocial components (Woolnough and Fielden, 2014).
The findings for gender and mentorship outcomes are mixed. One study found that male mentors are beneficial for women in the workplace; specifically, women with male mentors typically get more promotions and higher pay (Dreher and Cox, 1996). However, while a study of female students in economics found that female doctoral students with female faculty mentors took longer to graduate than did female doctoral students with male faculty mentors (Neumark and Gardecki, 1998), a later study found that female mentees with male mentor matches resulted in the women going on to a research position as their first job more often than male-male matches (Hilmer and Hilmer, 2007).
How Does Mentor-Mentee Matching on Social Identities Affect Mentorship?
Research is equivocal on the value of same-race and same-gender mentoring relationships. Mentees can benefit from mentoring relationships matched on both deep and surface levels (see Box 3-2). As noted in Chapter 1, effective mentorship is based on the ability of mentors and mentees to trust, share strengths with, identify with, and authentically engage with one another (Blake-Beard et al., 2011).
Some literature on UR STEM students and mentorship suggests that having mentors who are similar to mentees on key identities, such as race and gender, may produce benefits for UR students, especially in psychosocial support (Blake-Beard et al., 2011; Patton and Bondi, 2015).30 In addition, same-race and same-gender pairings had the potential to provide an understanding of shared experiences of being underrepresented in STEM spaces (Felder and Barker, 2013). Having a mentor who has been through similar experiences based on a shared identity also benefits mentees in terms of identification, developing interpersonal comfort, building trust, and setting expectations.31 Studies have also found that shared social identity in mentorship is more likely to engage the student holistically (Baker and Griffin, 2010; NASEM, 2017c; Pfund, 2016).
UR students in research training programs mentioned the value of seeing others like themselves (i.e., in race and gender) as a motivating factor in pursuing STEM advanced degrees (Hurtado et al., 2009).32 Same-race connections allow Black doctoral students to experience meaningful validation, affirmation, and success, which one study has shown to be crucial for completing their doctoral programs (Barker, 2011).33 Moreover,
30 In organizational research, demographic similarity between mentor and mentee has been linked to higher levels of mentees’ career support, and to mentors feeling a closer connection with their mentees (Ensher and Murphy, 1997).
32 When students who might otherwise feel ignored see themselves in and receive support and guidance from a similar individual who is a successful STEMM professional, it can help them to feel recognized and appears to strengthen science identity.
33 Such connections also served as a visual representation that confirmed the students’ participation in STEM programs.
in same-race and same-gender mentoring relationships, mentees witness firsthand and experience secondhand what their mentor does, thereby gaining a sense of self-efficacy and confidence that they too will succeed (Williams et al., 2016a).34
Some studies have found that many UR students want mentors of the same race and gender and who have life experiences similar to their own, including experiences pertaining to race, ethnicity, and gender (Blake-Beard et al., 2011; Williams et al., 2016a), and many seek these role models at minority-serving institutions (Hurtado et al., 2009; NASEM, 2019). One study focused on mentoring outcomes in STEMM found that an overwhelming majority of over 1,000 racially diverse undergraduate and graduate STEMM students surveyed felt it was important to have a mentor of the same race and gender (Blake-Beard et al., 2011). Respondents in same-race and same-gender mentoring relationships were more likely to report they had received more career and psychosocial support. However, there were no apparent effects of this greater amount of mentoring for outcomes such as increased grade point average, self-efficacy, or confidence about their fit in science (Blake-Beard et al., 2011). The participants, particularly UR students, felt it was important that mentors understand how students’ backgrounds could affect their professional careers. This suggests that while mentees may prefer social identity matching with their mentors, what is ultimately important is the mentor’s acknowledgment of the role of students’ social identities in their career development. Moreover, some workplace mentoring research indicates that mentors from a well-represented background can use their available social capital through the mentoring relationship to benefit the mentee’s career support and outcomes (Eby et al., 2013; Johnson and Smith, 2016), suggesting one potential benefit of cross-identity mentoring relationships.
While surface similarities may be important for some students, deep-level similarities such as having shared interests, values, and goals is also important for effective mentoring relationships, even across cultural differences. Mentors and mentees having deep-level similarities (Harrison et al., 1998) predicts interpersonal comfort, which in turn predicts psychosocial and career (instrumental and networking) support (Brunsma
34 This type of psychosocial support is commonly referred to as role modeling.
Opportunities to maximize matching along various demographics such as race are challenged by the scarcity of UR faculty in STEMM. For example, in 2015, of the 248,500 science and engineering faculty in the United States, 8,600 faculty were Black (3.5 percent of the total), 11,850 were Hispanic (5 percent), and 500 were Native American (less than 0.33 percent) (NCSES, 2017). When UR students struggle to find a faculty member of their race, gender, or sexual orientation, peer mentoring and near-peer mentoring may provide an alternative or additional option.
Another predictor of STEMM success for UR students was mentor and mentee “fit,” which is when the area in which the mentee needed support was an area in which the mentor could provide support (Baker and Griffin, 2010; Blake-Beard et al., 2011). For example, if mentees require more career support, it may be more effective for them to seek out individuals with more career-based social capital to assist them.
UNDERREPRESENTED SEXUAL- AND GENDER-MINORITY STUDENTS IN STEMM
Until recently, issues related to sexuality and gender have received little attention in STEMM (Yoder and Mattheis, 2016), and relatively few studies have explored sexual- and gender-minority identities in the STEMM fields (Cech and Waidzunas, 2011; Riley, 2008).35 Additionally, sexual orientation and gender identity may not be as visible as some other characteristics of UR students, such as race and gender. Sexual- and gender-minority students can decide not to disclose their orientation to colleagues, but this may result in feelings of invisibility, isolation, and rejection or hiding part of their identity. For example, students learn to take stock of the environment to manage their gay identity along with a strong engineering identity (HHMI, 2016). However, believing or actually needing to hide one’s identity can contribute to stress and negative mental health outcomes (Meyer, 1995; Pachankis, 2007) and create a strain on social relationships (Yoder and Mattheis, 2016), which may reduce workplace productivity even without active discrimination (Clair et al., 2005; Patridge et al., 2014).
35 The term sexual and gender minority is consistent with current language of U.S. federal agencies. See, for example, the Sexual and Gender Minority Research Office of the National Institutes of Health (more information is available at https://dpcpsi.nih.gov/sgmro; accessed August 17, 2019). For the purposes of this report, sexual- and gender-minority students include, but are not limited to, students with sexual orientation identities such as lesbian, gay, bisexual, queer, and asexual, as well as gender identities such as pre- and posttransition transgender, intersex, and nonbinary.
found that individuals in academia are less likely to know the kind of support their employers provide to sexual- and gender-minority employees (Yoder and Mattheis, 2016). Research has also found that sexual- and gender-minority faculty in STEMM fields with higher rates of women representation reported a higher degree of openness to sexual- and gender-minority needs (Yoder and Mattheis, 2016).
Students sometimes encounter silence or assumptions about their major as a “man’s field” that continues to marginalize both women and gay men (HHMI, 2016). While many departments are aware of sexual- and gender-minority rights, most do not understand the efforts needed to address the issues adequately. In an effort to improve the institutional climate regarding sexual and gender identity in STEMM, a sexual- and gender-minority physicists advocacy group created a Best Practices Guide that addresses areas such as using gender-neutral and inclusive language, inviting sexual- and gender-minority speakers to campus, and joining ally groups (Ackerman et al., 2018). Broad institutional support can help create a supportive environment in which faculty and students feel comfortable being “out” about sexual identity (Ackerman et al., 2018). As STEMM works to diversify its faculty and students, it is crucial to create an environment in which faculty and students can be out and to make this awareness part of the mentoring process for students in an environment that may be discriminatory toward the sexual- and gender-minority community. Further research has been called for on the role of out mentors and how they can help students who have self-selected to leave STEMM fields because of discomfort caused by intolerance (Yoder and Mattheis, 2016).
Similarly to other aspects of identity, sexual- and gender-minority students in STEMM face challenges that involve a disregard for gender and sexual identity owing to the high value placed on science and the scientific identity. There is a lack of understanding about efforts to create sustainable and equitable changes that allow sexual- and gender-minority students to feel comfortable being open. Ambient heterosexist harassment, often related to campus climate, has detrimental effects on both sexual minorities and heterosexual students’ psychological well-being and feeling comfortable on campus (Silverschanz et al., 2008).37 Sexual- and gender-minority individuals also face neglect or encounters with many discriminatory practices and policies, such as the refusal of institutions to provide gender-neutral restrooms. In addition, there is inconsistent protection for sexual orientation and gender identity in nondiscrimination laws by state.38 In diversifying STEMM, more support and research is required to improve mentorship practices for sexual- and gender-minority populations. For example, Safe Space training for mentors at the institutional, departmental, or unit levels and placards for faculty offices could indicate support
37Ambient heterosexist harassment is defined as “insensitive verbal and symbolic (but non-assaultive) behaviors that convey animosity toward non-heterosexuality” that “take place within the environment but are not directed at a specific target, such as the telling of [heterosexist] jokes that can be heard by anyone within earshot” (Silverschanz et al., 2008, p. 180).
38 For example, see https://www.aclu.org/files/pdfs/lgbt/discrim_map_bw.pdf; accessed September 20, 2019.
for sexual- and gender-minority students.39 Faculty who identify as sexual or gender minorities in particular may engage support for being out so they can mentor students who are also out but are leaving STEMM fields because of bad experiences.
UNDERREPRESENTED STUDENTS WITH DISABILITIES IN STEMM
The American Disabilities Association defines disability as “a physical or mental impairment that substantially limits one or more major life activities.”40 For legal purposes, this includes those who have documentation of an impairment, even if they are not registered as having a disability (Francis, 2018). Disabilities can be both readily discern-able (e.g., loss of limb, visual impairment) as well as significantly less so (e.g., learning challenges, mental health challenges). Students with disabilities enroll in undergraduate STEMM programs at rates approximately two percentage points lower than students who do not report any disabilities (Alexander and Hermann, 2016; NCSES, 2013). This does not, however, necessarily indicate a lack of interest in STEMM, because people with disabilities pursue STEMM degrees at the same rate as those without disabilities (Thurston et al., 2017).
One issue that has been identified pertains to the increase in dropout rates between high school and college, and again between undergraduate and graduate school (Booksh and Madsen, 2018). Many students with a disability struggle with going from a structured high school and family setting to a university setting with new freedoms and less structure. Students with disabilities typically have had individualized education programs or 504 plans and a support team of teachers, parents, and educational support staff in K–12 schooling, but in college the students are left largely to their own efforts to obtain accommodations (Kurth and Mellard, 2006). Colleges may offer disability services but not at the same level of integration and monitoring as K–12 schools. A study of 110 undergraduate students found that less than a quarter of students who have individualized education programs or 504 plans register with college disability services, and only 60 percent of those receive accommodations (Cawthon and Cole, 2010). If students with disabilities start to fall behind their peers, they are less likely to persist. Students with disabilities also report that a lack of support from the academic community creates a feeling of not belonging in a group and shame associated with the disability (Booksh and Madsen, 2018). Research with deaf/hard-of-hearing mentees indicates that effective mentorship practices may help to alleviate this (see Box 3-3).
Some students with disabilities may have received mathematics and science preparation in specialized programs in middle and high school that does not align with specific requirements stated for undergraduate mathematics and science courses (Lynch et al.,
40 See https://www.law.cornell.edu/uscode/text/42/chapter-126; accessed August 17, 2019.
2018). Faculty members, administrators, and staff may even show a lack of cooperation and understanding of the needs of students with disabilities. At least one study indicates there are often not enough adaptive aids, accessible spaces, and accommodations to adequately meet their needs in STEMM fields in particular (Moon et al., 2012). Another study has shown that there is little recruiting of students with disabilities into STEMM and a lack of methods to accurately measure the effectiveness of programming to help students with disabilities (Thurston et al., 2017).
Students with disabilities in STEMM may or may not require special accommodations to enable them to succeed in graduate education. Since disability can occur throughout one’s life, a better understanding of the onset of disability diagnosis and its influence on STEMM enrollment and degree persistence would benefit university disability services in providing developmentally specific supports in place for students who have recent disability diagnosis.41 Research has generated some lessons on facilitating the success of college students with disabilities. Having students with a disability use institutional disability services and existing resources allows faculty to focus on STEMM content, peer tutoring, lab communities, improved recruitment strategies, self-advocacy programs for students, professional development, and mentorship programs for students with disabilities in STEMM (Thurston et al., 2017). Research has also found that e-mentoring is an effective way to reach students with disabilities and improve persistence through self-advocacy and self-determination (Gregg et al., 2016).42
Mentorship is promoted for individuals with disabilities for the same reasons that it is promoted for other individuals. For individuals with a disability, mentorship appears to be particularly important when students go through transitions, such as from high school to college and from college to graduate school or to a job (Lindsay et al., 2016; Weir, 2004; Whelley et al., 2003; Wilson, 2003). A systematic review of mentorship programs designed to ease the transition from high school to university found that mentoring relationships for individuals with disabilities produce significant improvements in self-determination, empowerment, self-efficacy, and self-confidence or self-advocacy (Lindsay et al., 2016). Mentors of mentees with disabilities have also reported gaining benefits from their experiences with their mentees (Hillier et al., 2018; Stumbo et al., 2008, 2009).
In 2014, a multiauthored compilation and synthesis of programs and perspectives on fostering access to STEM careers among students with disabilities, entitled From College to Careers: Fostering Inclusion of Persons with Disabilities in STEM, was produced with support from the National Institutes of Health Director’s ARRA Pathfinder Award to Promote Diversity in the Scientific Workforce (Duerstock and Shingledecker, 2014). The compilation highlighted mentoring practices that included aspects of universal design for learning;43 a blend of in-person, virtual, and social media platforms to develop personal and professional networks; and accessible resources to support career aspirations in STEM.
A unique approach to mentoring that is described in this compilation is “developmental advising” that “combines aspects of academic advising, counseling, mentoring, and case management to provide students with a formalized single point-of-contact for support in pursuit of their educational and career goals” (Creamer and Creamer, 1994). As noted above, individuals with disabilities often face additional obstacles during transitions from high school to college, and this form of support with a strong, ongoing mentorship goes far beyond typical advising. However, this and any other forms of institutional support cannot replace the critical need for mentorship with science faculty, and access to legitimate research experiences, for students with disabilities, as for all other students. Like other mentorship interventions described in the compilation, short- and long-term interventions data for mentoring students with disabilities is lacking.
One group of people with disabilities who pursue STEMM majors are those diagnosed with a neurodiverse condition, such as autism spectrum disorder (ASD). Among students entering colleges with disabilities, those with ASD have the third-lowest rate of attending college, but a higher percentage of them select STEMM majors (White et
43 Universal design for learning, or UDL, is an approach to curricula and teaching methods that strives to be more inclusive than American with Disabilities Act guidelines.
al., 2011). It has been theorized that students with ASD are above average on creating systems, doing analysis, and understanding rule-based systems that help them excel academically in certain STEMM majors (Austin and Pisano, 2017). However, they score below average on emotional and social thinking, which can become a barrier to their success in college (White et al., 2011).
Students with ASD who have the academic skills and strengths to succeed often also have differences in sensory and executive functions and communicate in nontypical fashions that may result in problems of understanding and create unique challenges (Boutot and Myles, 2011). A critical component of postsecondary education is navigating the classroom environment and interactions with faculty and staff (Austin and Peña, 2017), and research has shown that students in general who interact more with faculty experience more satisfaction with their education, attain better grades, and have greater persistence to graduation (Harris et al., 2011). As a result, a faculty member’s interaction with a student with ASD plays an important role in that student’s success, with the perceived attitude toward providing accommodations for the student being a big factor in that success. Faculty members are legally required to provide “reasonable accommodations,” but students with ASD often do not follow through with registering at the university’s disability services or notifying faculty members of their needs (Austin and Peña, 2017). Faculty members are often aware of the needs of individuals who are blind or deaf, but more training is needed to make faculty members aware of the needs of people with ASD (Taylor, 2005).
Few articles have addressed faculty experiences with strategies for working with students with ASD. One group of investigators has outlined three strategies found to be effective, at least in the context of didactic instruction: minimizing classroom anxiety, improving executive functioning, and supporting critical thinking instruction (Shmulsky and Gobbo, 2013). The same investigators found that providing structure and giving attention to the classroom’s emotional climate were effective support strategies (Gobbo and Shmulsky, 2014). In the context of research experiences and research mentoring specifically, another study reported preliminary results from a pilot program of peer mentoring for university students on the autism spectrum. These results included gains in student’s self-reported measures of social support and general communication (Siew et al., 2017).
One study has reported results from a pilot undergraduate research program for engineering students with attention deficit hyperactivity disorder (Hain et al., 2018).44 Students in this program participated in extracurricular research projects that allowed them to work, interact, and connect with other neurodiverse students and use their intellectual strengths in a way that might be confined in the traditional engineering course environment. The study found that this intervention increased the participants’ interest
44 Attention deficit hyperactivity disorder, or ADHD, is sometimes considered a neurodiverse condition.
in engineering research, their sense of belonging in engineering, and their interest in pursuing graduate studies in engineering.
However, in contrast to the case for other identity groups or even including individuals with other disabilities, there is virtually no literature on empirical approaches, let alone theoretical constructs, for STEMM mentoring of students with ASD. With the expected growth of this segment of the postsecondary population, this appears to be a large and ripe area for research. Nonetheless, in light of the differences that define ASD, it is reasonable to surmise that mentoring strategies for ASD students may require significant modification from those used with other students in STEMM.
This page intentionally left blank.