Exploring the Relationship Between Faculty Mentor Engagement and African American STEM Persistence

This quantitative study explores factors associated with the persistence rates of African American students aspiring to major in STEM subjects and who ultimately complete undergraduate STEM degrees. The primary data source for this study is the Cooperative Institutional Research Program’s 2009 The Freshman Survey (TFS) and the 2013 College Senior Survey (CSS). The sample included 379 African American students who indicated on TFS that they intended to declare STEM as their major. The findings reveal that African American undergraduates are significantly more likely to persist in STEM majors with increased levels of faculty mentor engagement. The findings demonstrate the need for institutions to implement curricula that encourage meaningful faculty engagement to contribute to the creation of more inclusive academic environments for African American students in STEM.


Introduction
African American students graduate at significantly lower rates than their racial/ ethnic counterparts in science, technology, engineering, and mathematics (STEM) (Chang et al., 2014).Persistence in STEM is low among students (The National Academies of Science, Engineering and Mathematics 2016); however, a disproportionate number of African American students do not complete STEM programs compared to other groups (Lancaster and Xu 2017).Higher education stakeholders continue to seek solutions to address this disparity and achieve degree completion parity.Underrepresented minoritized students express interest in STEM at the same rates as their counterparts, yet their rate of completion of STEM degree programs is significantly lower (NASEM 2016;Hurtado et al., 2009;Eagan et al., 2011;Soldner et al., 2012).The shortage of African American STEM professionals has significant implications for the challenges facing our nation and for workforce demands.The President's Council of Advisors on Science and Technology (PCAST 2012:i) reports that the USA must produce a significant number of STEM graduates in order for the country to "retain its historical preeminence in science and technology." Addressing STEM degree completion disparities is a complex undertaking that requires a nuanced approach.More African American (AA) students are attending college than ever before; yet their completion of STEM programs continues to be at rates lower than their White and Asian counterparts (NASEM 2016).Between 1990 and 2013, AA enrollment into undergraduate programs increased by 5% (NCES 2016).While the increased enrollment of AA students in postsecondary education is promising, attrition from STEM degree programs is at higher rates than all other racial/ethnic student populations (Estrada et al., 2016).Lundberg and Schriner (2004: p. 562) explored the quality and frequency of faculty-student interactions and concluded that, "African American students experienced more differential treatment by faculty in academic settings than did Caucasian, Hispanic, or Asian students.The lower expectations held of AA students were conveyed by such behaviors as ignoring their participation, treating them stereotypically, and expressing impatience with their responses."AA students report that ineffectual formal relationships with faculty, unpredictable class offerings, and a lack of academic preparation contribute to the challenges they face in completing STEM degrees (Lancaster and Xu 2017).
Understanding what facilitates STEM degree completion helps to strengthen interventions created by researchers and practitioners focused on decreasing disparities.Research indicates that faculty/student relationships affect student satisfaction with their higher education experience (Guiffrida, 2005).However, the types of interactions that students have with instructors is key to understanding whether those interactions are beneficial to students.This paper examines the relationship between faculty mentor engagement and AA STEM persistence.

Key Factors Influencing the Impact of AA Student Outcomes
A diversified STEM workforce is necessary since minoritized students (MS) are more likely to examine problems faced by their communities than non-minoritized students (NMS) (Hurtado et al., 2008).The term "minoritized students" recognizes that students deemed to be "minority" are not inherently so.Rather, they are embedded in unjust and inequitable systems that consequently place them in the margins.Historically, pervasive literature has used the term "underrepresented minority students' to describe such students as those who identify as Hispanic or AA.However, 1 3 Journal of African American Studies (2023) 27:215-233 educational scholars have recently taken issue with the term 'minority' as it fails to address the actions that render an individual a 'minority" (Stewart, 2013).According to the National Science Board of the National Science Foundation (2021: p. 7), "Although Blacks or African Americans, Hispanics or Latinos, and American Indians or Alaska Natives represent 30% of the employed U.S. population, they are 23% of the STEM workforce due to underrepresentation of these groups among STEM workers with a bachelor's degree or higher."Of the AA students who entered STEM degree programs in 2004, only 13.4% completed their degrees within 4 years and 18.4% within 5 years (Higher Education Research Institute, 2010).In comparison, in 2004, White students had a 4-year STEM degree completion rate of 24.5% and a 5-year completion rate of 33% (Rohrbaugh & Corces, 2011).A number of factors have been shown to positively influence underrepresented students, including pre-college characteristics, STEM environment, sense of belonging, undergraduate research, and faculty mentors in undergraduate research programs (URPs).

Pre-college Characteristics
Pre-college characteristics, such as the amount of science and mathematics courses taken in high school or parental academic achievement, have been shown to have a positive correlation with STEM persistence (NASEM 2016;Dai & Cromley, 2014).However, the way mathematics is taught and the availability of mathematics courses in high school are potential barriers for STEM aspirants.The National Academies of Science, Engineering, and Mathematics (2016: p. 64) found that, "25 percent of the students who take calculus 1 at a research university receive a D or F or withdraw from the course, and another 23 percent receive a C." Student experience of mathematics curricula largely influences whether they perceive themselves as being capable of successfully completing their academic programs.

Institutional Environment
The institutional environment in which STEM is taught, and whether the culture of the institution promotes diversity and inclusion, influences the academic outcomes of STEM students.The National Academies of Sciences, Engineering, and Medicine (2016: p. 60) defines culture as the "shared patterns of norms, behaviors, and values of STEM disciplines that manifest themselves in the way courses are taught and the classroom is experienced."The National Academies of Science, Engineering and Mathematics (2016) has found that unwelcoming and hostile institutional cultures contribute to student isolation and withdrawal.One contributor to such an unwelcoming STEM environment is a stereotype threat, which is a burden that students feel when they perceive that they are viewed in congruence with group stereotypes (Hurtado et al., 2009;NASEM 2016).According to Appel and Kronberger (2012: p. 610), "stereotype threat is conceived as a state of psychological discomfort that is thought to arise when individuals are confronted with an evaluative situation, in which one's group is associated with a negative stereotype."Some studies have found stereotype threat to be an even more powerful predictor than academic preparation of persistence by MS students in STEM degree programs (Ben-Zeev et al., 2017).While science culture has been widely reported as unwelcoming to students of color, STEM academic environments may potentially be tempered by the presence of faculty members who demonstrate genuine care about the academic outcomes of students (Guiffrida, 2005).
The types of support that MS receive in college often determine whether they will persist in their academic programs (Chang et al., 2014).MS students report that postsecondary classroom environments often do not provide a healthy learning atmosphere (National Academies of Science, Engineering, and Medicine 2016).Negative racial experiences, including unwanted and unwarranted feedback on perceived social and cultural identity, often make AA students feel marginalized (Espinosa, 2011).

Sense of Belonging
In order to persist in STEM degree programs, AA students must feel a sense of belonging in their classrooms, and the feeling of being part of the overall science culture of their institution (Hurtado et al., 2007;Schlossberg, 1989).In addition to the interpersonal challenges apparent in science culture, certain structures perpetuate an atmosphere that may be perceived as being more concerned with "weeding students out of STEM majors" (Espinosa, 2011: p. 214) than with working to support and retain students once they have been admitted (Hurtado et al., 2008).Whether or not students experience their academic setting as supportive of their STEM potential influences their disposition as STEM aspirants and informs how they address obstacles and barriers along the way (National Academies of Science, Engineering, and Medicine 2016).According to NASEM (2016: p. 67), "academic climates that emphasize learning, mastery, and improvement in math and science, rather than inherent ability, can promote both performance and persistence in those subjects through positive effects on students' self-beliefs."Meaningful faculty mentor engagement may help to mitigate an unwelcoming science climate.

Undergraduate Research
ndergraduate research programs (URPs) have been found to positively influence MS STEM persistence.Such programs, Eagan et al., (2013a, b: p. 680) argue, typically include faculty mentors who, in addition to offering guidance on student research projects, provide "access to professional networks, new sources of information, and broader access to institutional resources."URPs also increase one-on-one interactions between faculty and students, thus bolstering the educational experience (Hurtado et al., 2009).Researchers have reported that URPs are positively related to the strength of a student's science identity (Eagan et al., 2013a, b;Hurtado et al., 2008), to the culture of the institution (Rodgers & Summers, 2008), and to the mitigation of the stereotype threat (Hurtado et al., 2008).Furthermore, they are positively correlated with the persistence in STEM of underrepresented minoritized students.While URPs help to mitigate MS STEM attrition, few studies focus specifically on whether faculty mentor engagement is positively correlated to AA STEM persistence.Toward that end, strategies that cater to a wider population of students are necessary to addressing STEM degree completion disparities.Exploring the relationship between the two could enhance interventions aimed at increasing AA STEM persistence.

Faculty Mentors
Faculty members strongly influence MS student decisions to remain in academic programs (Eagan et al., 2011;Gasiewski et al. 2012).McCoy et al., (2017: p. 658) found that "for undergraduate students, and Students of Color in particular, mentoring has been evidenced to improve students' transition to higher education, retention, GPA, self-efficacy, graduate student training, career selection, and socialization to academic and professional roles."Given the importance of faculty engagement in AA STEM persistence, this study defines mentoring as a series of behaviors that seek to encourage students, and enhance inclusion through meaningful engagement, in STEM.
Faculty mentor engagement encompasses more than simple interactions between faculty members and AA students and includes those faculty engagement behaviors that positively influence AA STEM degree completion.The mentoring experience includes students receiving advice about academics, critiques of their work, and graduate school or career preparation.The underlying key to faculty mentor engagement is a genuine concern about the successful academic or professional outcomes of AA students in STEM, which may span beyond what is typically outlined in faculty job descriptions (Guiffrida, 2005).

Faculty Mentors in URPs
URPs are positively correlated with MS STEM persistence due to faculty interactions, mentoring opportunities (Eagan et al., 2013a, b;Jones et al., 2010), and academic support (Hurtado et al., 2009).Moreover, MS science identity has been found to be strengthened by URPs (Hurtado et al., 2009).While URPs vary across institutions, a fundamental component of all URPs is engagement with faculty.Faculty mentors, in URPs, positively contribute to MS STEM persistence; however, Kuh and Hu (2001) have found that faculty are least likely to communicate with students by way of URPs.This finding is particularly concerning given that URPs are positively correlated with STEM persistence among MS students.If engagement with URPs is positively related to AA STEM persistence, and there is limited availability of URPs, then postsecondary institutions could further support AA students in STEM by utilizing the information presented in this study.

Theoretical and Conceptual Framework
This study is guided by the research question: What is the relationship between faculty mentor engagement and African American student STEM persistence?The Input, Environment, Output model (IEO) was used to emphasize the importance of examining campus environments and, more specifically, policies, practices, and procedures that positively or negatively affect the academic performance of AA STEM students (Astin, 1970;Pascarella & Terenzini, 2005).One example of policies and practices that may contribute to whether AA students persist in stem is the placing of "gatekeeper" classes in the first two years of a STEM degree while not simultaneously focusing on faculty diversity and inclusive pedagogy (Hurtado et al., 2009).Schlossberg's (1989) marginality and mattering framework expands the IEO model by amplifying the experiences of students in the unexamined margins.Given that environment has some bearing on AA students' decisions to continue with their programs of study, understanding the components of mattering, including attention, importance, ego-extension, dependence, and appreciation (Schlossberg, 1989), helps to contextualize the issue.
The role of self-perception on academic outcomes is also significant (Dai & Cromley, 2014;Eagan et al., 2013a, b).Despite what may be perceived as an unwelcoming science culture, there are those AA STEM students who persist in degree programs.This study was guided by Harper's asset-based approach to explore which educational interventions have been proven to work.Harper's (2010) anti-deficit model encourages a positive approach to AA STEM persistence, rather than focusing on what is perceived as not working.In this study, we used a strength-based lens, in part, by exploring the faculty interaction construct which provides faculty and students with specific academic ways of being that positively influence AA STEM persistence.The variables used for this study were also informed by the conceptual models of marginality and mattering, and by social and cultural capital, which informed factors associated with campus culture and access to information to advance in the science profession.Moreover, faculty members are positioned to augment the social capital of AA STEM aspirants; the persistence of AA students in STEM is positively influenced when faculty genuinely demonstrate care for the well-being of those students (Guiffrida, 2005).
Drawing on these theories, we established a conceptual framework that hypothesized that because majoring in STEM can be isolating for all students, and AA students in particular, faculty mentor engagement may lead to higher STEM persistence.Faculty mentors are the mechanism by which social and cultural capital are passed along to AA STEM students, as they work to demystify what it means to be a scientist, thus strengthening science identity and self-efficacy among students (Chang et al., 2014).As noted, college and university campus environments may be perceived as unwelcoming to AA STEM aspirants.Therefore, it is essential to explore additional strategies that will support a wider population of students in STEM.First-time full-time AA students in STEM at predominantly White institutions are the focus population of mentoring in this study.

Data Source
The primary data for this study came from the Higher Education Research Institute (HERI) at the University of California, Los Angeles.HERI houses the Cooperative Institutional Research Program (CIRP) which facilitates longitudinal surveys throughout the USA to understand student experiences in postsecondary education.The Freshmen Survey (TFS) is administered to students in their first year and the College Senior Survey (CSS) is administered to the same students in their final undergraduate year.The TFS and CSS enabled us to examine whether AA students in STEM are more likely to persist in academic programs when they engage in meaningful faculty mentor interactions.We define persistence as whether AA students indicated an interest in STEM on the TFS and then selected a STEM major on the CSS.Students who expressed a clear interest to major in STEM, but then switched to majors outside of STEM, were identified as students who did not persist in STEM degree programs.

Cooperative Institutional Research Program's TFS and CSS Survey
The CIRP survey is operationally used to (a) simultaneously collect data from students at different US institutions, and (b) collect longitudinal data to measure change (Astin, 1972).The Freshman Survey (TFS) is administered when students are accepted into postsecondary education as freshmen, and the follow-up College Senior Survey (CSS) is administered 4 years later in their senior year.Participating institutions are representative of 2-year and 4-year institutions.They self-select to participate in the surveys and pay to access the survey instruments and associated reports (Astin, 1972).Our interest in the CIRP is specifically focused on TFS and CSS respondents that identify as AA.The study focused on questions in the TFS and CSS survey that involve campus culture, faculty interactions, and how students navigate their academic studies.The CSS provides longitudinal data on the impact of the college environment on the academic performance of AA STEM students (Franke et al., 2010;Higher Education Research Institute, 2016).

Variables/Constructs
The faculty mentor construct used for this analysis consisted of thirteen distinct ways of being that help to inform the way classes are conduced, and the creation of meaningful mentoring relationships for AA students in STEM.The faculty mentor engagement construct was composed of responses to CIRP questions that ask about whether faculty: • Provide encouragement to pursue graduate/professional study, • Provide an opportunity to work on a research project, • Give advice and guidance about a student's educational program, • Provide emotional support and encouragement, • Provide a letter of recommendation, • Give honest feedback about a student's skills and abilities, • Help improve student study skills, • Give feedback on a student's academic work (outside of grades), • Are intellectually challenging and stimulating, • Provide an opportunity to discuss coursework outside of class, • Help students achieve their professional goals, • Provide an opportunity to apply classroom learning to "real-life" issues • Provide students with an opportunity to publish their work.
Each variable associated with the research question was then operationalized for each statistical test.The faculty mentor engagement construct variable was created by HERI using exploratory factor analysis of thirteen survey questions.In addition, three categories of high, medium, and low were created for each construct and students were assigned a category based on the rescaling of their scores.Scores of 0.5 standard deviations above the mean or higher were coded as "high"; scores within 0.5 standard deviations of the mean were coded as "medium"; and scores of 0.5 standard deviations below the mean or lower were coded as "low" (Frankie et al. 2010: p. 13).In this way, faculty mentoring engagement was used as a continuous and dichotomous variable to answer the study's research question.
We used a CIRP definition of STEM comparable to National Science Foundation approved STEM fields (National Science Foundation 2017).The majors considered to be STEM in this study were biology, biochemistry or biophysics, botany, environmental science, marine (life) science, microbiology or bacteriology, zoology, other biological sciences, aeronautical or astronautical engineering, civil engineering, chemical engineering, computing engineering, electrical or electronic engineering, mechanical engineering, other engineering, astronomy, atmospheric science (including meteorology), chemistry, earth science, marine science (including oceanography), mathematics, physics, statistics, and other physical sciences.
As noted in Table 1, the research question explored the relationship between faculty mentor engagement and AA STEM student persistence.Each regression analysis consisted of different combinations of control variables that account for factors that may influence the relationship between faculty mentor engagement and STEM persistence.Table 1 includes faculty mentor engagement as the independent variable and AA STEM persistence as the dependent variable.In the second model, we controlled for student characteristics and institution type.

Research Design
We answered the research question employing linear probability models and logit regressions.Both statistical tests were used to investigate whether the raw outputs demonstrate the same direction and significance for the outcome associated with each statistical analysis.Although the dichotomous nature of the outcomes of 1 3 Journal of African American Studies (2023) 27:215-233 interest made logit regressions the most appropriate statistical model, we focused our discussion of the results on the linear probability models as our interpretation of them as changes in outcome probability were more intuitive than the logit coefficients.However, we also noted whether the two approaches differed in direction or significance.
The logit model takes the form of

The linear probability model takes the form of
Where Y is the dependent variable of interest, X is the independent variable of interest, and X 2 − X n are the control variables as described in Table 1.

Power Analysis
A power analysis was conducted to determine a minimum meaningful effect size for analysis.The alpha level used to conduct the power analysis was 0.05, and the number of predictors was one (i.e., the research question).The predictor for the research question was faculty mentor engagement.Power analysis revealed that with a power of 0.80 and one predictor, the sample had to be approximately 150 students.The effect size for the population of students examined in this study was ≈ 0.053.The sample size for this study was 379 AA STEM students.

Limitations
The TFS captures the intended majors of newly admitted AA students at predominantly White institutions (PWIs), and not the majors they actually declared.The implication of capturing the desired majors of participants, and not their actual declared majors, is that the results do not reflect the actual majors of all survey participants.While the results accurately reflect the relationships between independent and dependent variables, some students may have chosen to major in non-STEM fields.The use of the intended majors of STEM degree aspirants is in direct alignment with similar research on MS STEM persistence (Chang et al., 2014;Eagan et al., 2013a, b).
AA students who began their postsecondary education at 2-year colleges and then transferred into STEM programs at PWIs were not included in the sample population used for this study.Data could not be obtained from those students who completed the TFS but who did not complete the CSS in their senior year.Knowing whether those students remained in STEM, transferred to other programs, or left the university would provide researchers with a more comprehensive understanding of AA student attrition from STEM degree programs.Further, the data do not Journal of African American Studies ( 2023) 27:215-233 include whether faculty mentors were in STEM roles or the proportion of diversity of STEM faculty at the PWIs in question.This impedes understanding of which faculty members mentor AA STEM students, and whether faculty members who exude those characteristics are in STEM.The extent to which the findings can be applied to AA STEM students is also limited as the population does not include AA students that are not first-time, full-time, AA STEM aspirants.Moreover, students who transferred between 4-year institutions were not included in the analysis.

Results
This study sought to explore whether faculty mentor engagement positively influences the academic outcomes of AA STEM aspirants.The data used for this study were collected and managed by CIRP at the University of California, Los Angeles, using the TFS and the CSS instruments."Positively correlated" results imply tests that were statistically significant in the positive direction.Tables 2,  3, and 4 in the "Results" section reflect the non-standardized coefficients of each statistical test.For statistically significant non-standardized coefficients, the level of significance may be viewed via the level indicator at the base of each linear probability model statistical table.
Approximately 75.5% of students in the dataset attended 4-year liberal arts colleges and 24.5% attended research universities.Colleges, in this case, are generally defined by having only one "school," while universities are institutions with multiple "schools" (e.g., education, liberal arts, science).Of the institutions represented in this study, 91% were private and 9% were public.Students identifying as female accounted for 63% of survey respondents and students identifying as male accounted for 36%.
Of the students who persisted in STEM, 29.65% majored in Psychology, 18.09% in Biology, and 6.03% each in Chemistry and Computer Science (see Table 2).Of the AA students who completed both surveys, 24% did not persist in STEM.
Table 3 presents a list of variables used to illustrate the percentage breakdown across AA students who persisted and did not persist in STEM degree programs.As shown in Table 4, those AA students who persisted in STEM had higher levels of faculty mentor engagement (50.36%).
Table 4 presents a linear probability analysis with STEM persistence as the dependent variable and faculty interaction as the primary independent variable.We examined the extent to which the relationship between faculty mentor Income between $150,000 and 250,000 9.1 6.9 1 3 Journal of African American Studies (2023) 27:215-233 engagement and AA STEM persistence changed with the addition of other variables.Model 1 had no control variables as we were only interested in exploring the correlation between faculty interaction and STEM persistence.In this simple model, there is a positive correlation between STEM persistence and faculty mentor engagement.In Table 4, faculty mentor engagement continued to reflect a positive relationship with AA STEM persistence even with the addition of student characteristics, including academic preparation, parental academic achievement, first-year generation status, and institution type.However, the relationship between faculty mentor engagement and STEM persistence weakened to the point of non-significance with the addition of annual family income.This may be a consequence of the small sample size and adding multiple variables into the model.As shown in Table 4, while the relationship between faculty mentor engagement and STEM persistence weakens with the addition of the control variables, the relationship remains marginally significant.The raw output magnitude of the faculty interaction coefficient did not change much when student characteristics, institutional context, and faculty diversity were added in models 2, 3, and 4.

Summary
This study focused on testing the relationships between faculty mentor engagement and AA STEM persistence.The variables that underpin this study emerged from a comprehensive critical review of the literature which guided the selection of the theoretical and conceptual frameworks.Linear probability models and logit regressions were conducted to examine the relationships between the independent and dependent variables to ensure that the statistical outputs were similar with respect to raw output magnitude and direction.
Based on the results from the linear probability and logit regressions, faculty mentor engagement was found to have a positive correlation with AA STEM persistence (p < 0.05).This finding indicates that AA STEM students engaged in faculty mentorship have an increased chance of persisting in their degree programs compared to those who are not.This finding is helpful for students and faculty both in and outside of URPs to enhance their work and interactions.For institutions with URPs, examining the objectives of faculty-student mentor engagement, reminiscent of those listed in this study, can ensure effective interactions rooted in practices that have been proven to positively influence AA STEM persistence.For students and faculty in institutions that do not have URPs, this finding can guide interactions while being mindful that, in order for interactions to be useful, they must be meaningful.Our main finding is that AA students with more faculty interaction have an increased likelihood of persisting in STEM degree programs.

Discussion
STEM degree completion requires that AA students navigate a complex web of internal and external factors that ultimately determine whether they will persist in their programs of study.It is incumbent upon higher education researchers and practitioners to consider the multilayered complexities of STEM persistence when creating interventions.Based on best practices in STEM persistence research, higher education professionals have launched formal URPs to mitigate MS STEM degree 1 3 Journal of African American Studies (2023) 27:215-233 departure.These usually consist of faculty interactions, research opportunities, and avenues for students to present their research findings (Hurtado et al., 2009).URPs are widely regarded as having the ability to strengthen MS science identity, selfefficacy, and sense of belonging, all of which positively influence the likelihood of STEM degree completion (Hurtado et al., 2009).Additional intervention strategies are necessary to increase the rate of STEM degree completion for students from underrepresented communities.Therefore, examining the types of interactions between faculty and AA STEM students can help to bolster academic support for a wider range of students.

Interpretation of the Results
Our research question was developed to address a gap in the literature on AA STEM persistence, and to bolster support for AA students in STEM.Academic outcomes clearly reflect a disproportionate amount of attrition among AA students in STEM degree programs.Our research found that particular types of faculty mentor engagement is positively correlated with AA STEM persistence, and that students who were one standard deviation higher in faculty engagement were 3.2% more likely to persist in STEM than others, when socioeconomic status was not included in the model.To be clear, faculty mentor engagement is positively related to STEM persistence, but this relationship, at least in part, is driven by family annual income.It appears that students who reported a higher annual family income may be more likely to interact with faculty and persist in STEM degree programs.This may be due to the sharing of social and cultural capital with regard to the navigation of academic space; however, more research is needed to confirm this.

Theoretical Recommendations
The theories that framed this study align well with the research findings.Postsecondary science culture is often characterized as unwelcoming, rigorous, and competitive, leading to AA STEM students feeling marginalized (Cole & Espinoza, 2008;Espinosa, 2011;Hurtado et al., 2008Hurtado et al., , 2009; National Academies of Sciences, Engineering, and Mathematics 2016).We hypothesized that students who engage with faculty mentors would have higher levels of STEM persistence because of the role played by faculty in helping AA students navigate the science and engineering culture (Hurtado et al., 2009;Jones et al., 2010;McGee, 2016).It appears that faculty members may enhance feelings of mattering (Hurtado et al., 2009;Schlossberg, 1989) among AA students in STEM by being available to answer questions (attention), providing emotional support and encouragement (importance), inviting participation in undergraduate research programs (ego-extension), facilitating opportunities to publish (dependence), and celebrating academic outcomes (appreciation).

Recommendations to Practice
Faculty who encourage AA STEM students to attend graduate school, provide research opportunities, give feedback, provide emotional support, and give advice on how to navigate the program are positively correlated with AA STEM persistence.It is understandable that faculty may feel stretched due to their teaching, research, and service responsibilities; however, considering how mentoring for AA students might be included in STEM curricula could be one way to leverage time and resources to increase AA STEM persistence (Eagan et al., 2011).
Leaders within higher education, such as chancellors, provosts, and academic deans, play a critical role in fostering environments that consider the academic experience of all students.Their diversity and inclusion leadership help to foster academic spaces that consider the academic success of everyone.Further, on-campus student-focused units, such as admissions, graduate divisions, counseling/advising, multicultural centers, writing centers, or associated student centers, should consider how the findings of this study might aid in the creation of inclusive on-campus spaces.Collecting data to ascertain who seeks and receives such services, and who does not, could reveal important information about why some students feel welcome and others do not within particular academic environments.Additionally, many of the items comprising the faculty mentor engagement construct could be used to inform academic programming such as the creation of programs on campus that focus on providing resources to AA students.

Conclusion
There is a disparity in STEM degree completion between AA students and their racial/ethnic counterparts.STEM degree attainment occurs by way of a complex pathway of opportunities and barriers that students must learn to negotiate in order to succeed.Higher education stakeholders must consider the multiple layers associated with AA STEM degree completion when creating strategies to support all students in their academic pursuits.The impetus for this study was to identify ways in which predominantly White institutions can further support AA students in STEM, as attrition from STEM programs is higher among AA students than among their peers.The findings of this study are encouraging as they help to inform scholars about strategies to strengthen academic outcomes for students who have already expressed an interest in STEM.The results of this study reveal that the items that comprise faculty mentor engagement in this study are positively related to AA STEM persistence.
The issue of STEM degree completion affects everyone.Excluding students from any sector of the economy, due to either internal or external factors, hampers the innovation, scientific solutions, and technological advancement necessary to maintain competitiveness on the national and international stage.STEM professionals must reflect the diverse needs of the USA.AA STEM degree completion must be 1 3 Journal of African American Studies (2023) 27:215-233 addressed in a layered and nuanced way to ensure the inclusivity of institutional culture for all STEM students.
Postsecondary faculty, staff, and administrators can use the findings in this study to ensure that campus environments are inclusive, and that all students, regardless of their racial/ethnic backgrounds, can thrive in STEM.In addition, understanding the role of faculty mentor engagement in AA STEM persistence informs higher education stakeholders about practical retention efforts that can strengthen support for greater diversity in STEM.This study has found that faculty mentor engagement is positively correlated with AA STEM persistence; therefore, researchers and practitioners may consider creating more opportunities to facilitate faculty-student interactions and ensure that faculty are aware of, and incorporate, engagement practices that have a positive influence on AA STEM persistence.

Table 1
Research question, theory, and operationalization of the variable

Table 3
Average score of AA

Table 4
OLS models for faculty mentor engagement and AA STEM persistence