Attending to STEM education in servingness at Hispanic-serving institutions: a systematic review of more than a decade of scholarship

Background, context, and purpose of the study Enrolling over 60% of all Latinx undergraduate students, Hispanic‑serving institutions (HSIs) are poised to play a critical role in diversifying and strengthening Science, Tech‑ nology, Engineering, and Mathematics (STEM) education and the STEM workforce. However, how HSIs serve STEM students is not well understood. Accordingly, guided by Garcia et al. (Review of Educational Research 89:5–745, 2019) multidimensional servingness framework, we conducted a systematic review of the research on STEM education within the HSI context. By attending to STEM education in conversations around how HSIs may serve Latinx students and their campus communities, our ultimate aim is to improve STEM education particularly at HSIs and advance STEM servingness more broadly. Results, main findings Through our systematic review of STEM education research at HSIs, we identified (under) studied components of servingness and gaps within this literature base. Specifically, among the 128 qualifying articles, nearly two‑thirds focused on student outcomes but overlooked institutions’ organizational context, rais‑ ing questions about the effect(iveness) of the studied interventions. Additionally, we identified three thematic gaps in this literature: ghosting the HSI context (i.e., relying on HSIs as research sites without considering the unique HSI context); ghosting Latinx culture (i.e., decentering Latinx students and the Latinx community’s sociocultural aspects and assets), and ghosting people and places (i.e., under‑examining certain student populations like Latino men in STEM and places like Hispanic‑serving community colleges). Ultimately, our study extends the field’s understanding of servingness by attending to STEM education within the context of HSI institutions. Conclusions, brief summary, and potential implications By systematically reviewing studies on STEM education at HSIs, we identified (under)studied components of servingness and patterned gaps within this literature. In doing so, we highlight opportunities to advance STEM servingness at HSIs through future research, policy, and practice. Col‑ lectively, these avenues hold the promise of improving STEM education and diversifying the STEM workforce.

Originally envisioned and built to serve white1 society, namely elite white men, many colleges and universities participated in the de facto exclusion of Black and Brown communities for more than 100 years in the United States.Even with the landmark Brown vs. Board of Education civil rights ruling that outlawed state-sanctioned segregation, de jure forms of exclusion still reign across much of U.S. higher education.Such exclusionary practices alongside unwelcoming, hostile, and chilly campus climates historically have stymied students of color's access to and success in higher education.Troublingly, innumerable studies continue to provide compelling evidence of the myriad ways colleges and universities still underserve and push out Black, Indigenous, and people of color (BIPOC) students.
Although an exclusionary ethos pervades U.S. higher education, scholars have especially implicated Science, Technology, Engineering, and Mathematics (STEM) fields in curtailing the full participation of BIPOC communities and reinforcing social stratification (Benjamin, 2019;Carter et al., 2016;McGee, 2020).Indeed, the literature has repeatedly called out the lack of racial/ethnic and gender diversity in STEM.Despite nationwide efforts and investment, two-thirds of the science and engineering workforce remains white (National Science Foundation [NSF], 2017).What's more, the literature emphasizes STEM fields' ongoing complicity in excluding BIPOC students and hampering their persistence and success in these areas.In particular, multiple studies have addressed the concerning underrepresentation of Latinxs2 in STEM majors and careers, despite their growing presence among the college-going population and among college graduates in the United States (Martinez et al., 2017;Rincón et al., 2020;Student Research Foundation, 2019).Illustratively, although Latinxs represented 36% of the total college-going population in 2018 (National Center for Education Statistics, n.d.), they accounted for only 12% of STEM graduates and 8% of all STEM professionals in the U.S. workforce in 2018 (Fry et al., 2021).
Enrolling over 60% of all Latinx undergraduate students in the United States today (Excelencia in Education, 2023a) and producing 40% of Latinx STEM graduates in 2010 (Núñez et al., 2015), Hispanic-serving institutions (HSIs)3 are slated to play a critical role in diversifying and strengthening STEM education and the STEM workforce (Jackson & Rudin, 2019; National Academies of Sciences, Engineering, & Medicine [NASEM], 2019).Recognizing this opportunity, various federal funding agencies, such as the NSF, the Department of Education (ED), and the U.S. Department of Agriculture (USDA), have invested in HSIs to ameliorate barriers in STEM education and increase the number of Latinx STEM graduates.
Despite such concerted public investment and HSIs' major role in educating racially/ethnically minoritized STEM students, limited research considers HSIs' transformative potential for advancing STEM education.Hence, drawing from Garcia et al. (2019) multidimensional framework of servingness, we conducted a systematic review of the literature on STEM education at HSIs to understand how scholars have approached or conceptualized serving racially/ethnically minoritized STEM students, especially Latinx STEM undergraduates, at HSIs.Specifically, we reviewed peer-reviewed scholarship that analyzed empirical data published from 2011 to 2022, guided by the following research questions: 1. What components of servingness has STEM education research (under)studied?2. What are the gaps in the research on STEM education at HSIs relevant to serving Latinx STEM undergraduates?
By pursuing these questions and explicitly attending to STEM education, this study is poised to contribute to HSI scholarship, particularly the growing body of work on servingness.This systemic review also offers faculty, researchers, and practitioners in STEM fields a valuable understanding of the literature on STEM education within the context of HSIs.Ultimately, by weaving together STEM education and HSI research, this study foregrounds possibilities to better serve Latinx and other BIPOC STEM students at HSIs and beyond.

Literature review
To set up this study, we first provide background on HSIs-their emergence and distinctive characteristics.Afterward, we discuss research on endemic issues in STEM education, emphasizing barriers to access and persistence for minoritized students in these fields.As part of this review, we also fold in relevant literature specific to HSIs.We then highlight the contributions of HSIs to STEM education.

The emergence and distinctive characteristics of Hispanic-serving institutions
Heeding Latinx education advocates' calls for funding parity, Congress federally recognized HSIs-and their pivotal role in the postsecondary education of Latinx students-in 1992 (Valdez, 2015).Over the last 30 years, this group of 2-year and 4-year institutions has grown in number and spread geographically.As of 2021-2022 estimates, there are 571 HSIs across 28 states, the District of Columbia, and Puerto Rico (Excelencia in Education, 2023a).In other words, approximately one out of every six colleges and universities in the United States enrolls the needed threshold percentage of Latinx and Pell-eligible undergraduate students (25% and 50%, respectively).
HSIs are enrollment-dependent institutions and are remarkably diverse as a population, from institutions focused on awarding associate degrees to those more focused on research (Núñez et al., 2016).Specifically, the most recent estimates indicate that 40% of HSIs are 2-year public colleges while 59% are 4-year institutions, with an almost even split between public and private (Excelencia in Education, 2023a).While recognizing this population's heterogeneity is crucial, particularly when considering servingness, this group loosely shares other core characteristics, which merit noting.In particular, by legislative definition, HSIs must operate with low core expenses.As expected then, many HSIs, albeit not all, have pressing resource constraints, running with minimal endowment holdings and limited access to additional revenue streams (Aguilar-Smith, 2024).At the same time, they shoulder the lion's share of the work of educating racially/ethnically minoritized and cash-strapped students.In short, these generally under-resourced institutions serve an outsize share of students who often have the greatest needs.
Despite these colleges and universities' unparalleled role in providing such students access to higher education, particularly Latinx students, with few exceptions, HSIs were not created with the explicit mission to serve Latinx communities the way Historically Black Colleges and Universities and Tribal Colleges and Universities were intentionally built for Black and Indigenous students, respectively (Laden, 2004).In other words, most HSIs are historically white institutions, becoming HSIs in response to racial/ethnic demographic changes in the United States.Thus, many HSIs still mirror predominantly white institutions (PWIs) in their forms and functions, meaning their leadership, faculty/staff composition, curricula, and programs often do not reflect the needs or values of Latinx communities (Garcia, 2015).Rather, many HSIs still uphold whiteness and white dominant standards that (re)produce harm, especially for BIPOC students (Garcia, 2015;Scott et al., 2022).

Endemic gender and race-based issues of STEM education
Research on STEM education extensively documents gender and race-based issues, which undermine women and BIPOC students' access to and persistence in STEM.As far as gender-based barriers, research has found that women are less inclined to choose STEM majors, and less likely to persist, because of the unwelcoming or "chilly" environment in many STEM disciplines across the world (Lee et al., 2021;OECD, 2018;Ro et al., 2021).To put it another way, women are not less interested or proficient in STEM; they are not "bad" at math (Reardon et al., 2019).Instead, these environments, particularly their domineering male-centered cultures, both deter women from entering these fields and push women out of these fields and careers (Archer et al., 2017;Lingyu et al., 2021;Wang & Degol, 2017).Additionally, given the limited gender representation in STEM, particularly in engineering fields (Ramirez & Kwak, 2015), women students, like BIPOC students, have voiced concern about the absence of STEM role models and mentors (Chesler & Chesler, 2002;Dennehy & Dasgupta, 2017;Young et al., 2013).Together, these issues magnify women's sense of "otherness" and isolation in STEM and result in their subpar support in these fields and careers (Archer et al., 2017;Atkins et al., 2020;Good et al., 2012).
Turning to race-based issues in STEM, BIPOC students, particularly from socially and economically disadvantaged communities, often face obstacles in taking advanced math and science courses required for entry into STEM majors (Riegle-Crumb & Grodsky, 2010;Smith et al., 2016).More specifically, the entrenched racial/ethnic and subsequent resource segregation of K-12 schooling in the United States limits many BIPOC students' opportunity to enroll in crucial prerequisite courses, thus curtailing their access to STEM majors (Bottia et al., 2015;Dalton et al., 2007).
Meanwhile, the literature on BIPOC students, particularly Latinx STEM students, is chock-full of studies and stories of acculturation stress, cultural incongruence, marginalization, and dissatisfaction.In particular, Latinx STEM students often confront chilly and unwelcoming environments that (un)intentionally uphold racism, sexism, and classism.These environments often center competition and foster a "weed-out" culture rather than the values of collaboration, interdependence, and collectiveness often central to Latinx culture (Herrera et al., 2022;López et al., 2019;McGee, 2016;Rendón et al., 2020;Rincón et al., 2020).This cultural disconnect is taxing, making Latinx STEM students' transition to and success in college all the more difficult (Núñez et al., 2021;Preuss et al., 2021;Sáenz & Ponjuán, 2011).Relatedly, as touched on earlier, BIPOC students, including Latinx students, also enter STEM classrooms and labs rarely seeing faculty similarly identified.The limited racial/ethnic representation among STEM faculty and the workforce not only likely exacerbates this cultural incongruence, but also leaves BIPOC students in these areas less able to see themselves as scientists, engineers, or mathematicians.
Importantly, designation as an HSI does not negate many of these persistent gender and raced-based issues within STEM education.Indeed, a few studies have highlighted difficulties due to the lack of diversity and understanding of Latinx culture and values in STEM, specifically for Latina STEM students at HSIs (e.g., Contreras Aguirre et al., 2020, 2021;Gonzalez et al., 2021;Herrera et al., 2022).For example, in their case study on Latina STEM undergraduate students' persistence and success at HSIs, Contreras Aguirre et al. ( 2020) participants noted the "overwhelming presence of White men" (p.815) and how this gender imbalance left them feeling the need to defend their place and intelligence to both their men peers and professors.Additionally, Gonzalez et al. (2021) noted the isolation Latina women students at HSIs often feel in their STEM courses and co-curricular activities.Despite all of these ongoing challenges within STEM education, HSIs meaningfully contribute to STEM education and the STEM workforce.

Hispanic-serving institutions' contributions to STEM education and careers
As previewed earlier, HSIs serve as an entry point to higher education for most Latinx undergraduates and many other racially/ethnically minoritized groups, including American Indian, Asian, and Black students (Núñez et al., 2015).Furthermore, research suggests that this group of nearly 600 postsecondary institutions, especially Hispanic-serving community colleges (HSCCs), provides pathways to STEM fields for BIPOC students (Crisp et al., 2009;Herrera & Rodriguez-Operana, 2020;Zamani-Gallaher et al., 2019).For instance, in their descriptive study of national trends on STEM pathways through HSIs, Herrera and Rodriguez-Operana (2020) indicated that 81% of Latinx community college students in STEM fields attend either an HSI or emerging HSI (eHSI) and that 4-year HSIs/eHSIs educate nearly 60% of Latinx STEM students.Additionally, HSIs enroll more women STEM students than non-HSIs (Herrera & Rodriguez-Operana, 2020)-a noteworthy contribution considering the persistent underrepresentation of women in STEM fields and careers.
In addition to offering STEM educational opportunities for students traditionally underrepresented in STEM fields (i.e., racially/ethnically minoritized students and women students), HSIs also contribute to diversifying the STEM industry and workforce, collectively graduating thousands of STEM degree-holders every year.Regarding STEM attainment rates, HSCCs and their 4-year counterparts graduate approximately 35% and 62% of all STEM majors, respectively.Additionally, 32% and 68% of Latinx STEM students at HSCCs and 4-year HSIs, respectively, graduate within six years (Herrera & Rodriguez-Operana, 2020).Altogether, HSIs, especially HSCCs, are crucial to advancing equity within STEM fields and careers.
Beyond these success and performance measures, increasingly, research supplies evidence of HSIs contributing to STEM education through change-making faculty members and asset-based programming for Latinx students.In terms of the former, Bensimon et al. (2019) found that Latinx STEM professors serve as recruiters, develop skill-building programs, integrate Latinx students into STEM networks, and broker support for Latinx students within the university and across systems.Meanwhile, in terms of the latter, Núñez et al. (2021), for example, showed how HSIs (within the Computing Alliance of Hispanic-Serving Institutions; [CAHSI]) developed talent and increased Latinx attainment rates in computing fields by centering Latinx students' cultural assets.More specifically, they found that CAHSI members weave the values of confianza (trust), respeto (respect), and familismo (familism, a commitment and sense of loyalty to one's family)-integral to the Latinx community-across their efforts.Their study showed how a segment of HSIs foster student validation and enact servingness.In sum, HSIs show promise in diversifying STEM along racial/ethnic and gender lines, supporting the persistence and overall success of racially/ethnically minoritized students and women of color in STEM, and ultimately advancing servingness.It is this multidimensional concept (i.e., servingness) that we turn to next when describing the theoretical framework for this study.

Theoretical framework
We drew on Garcia et al. 's (2019) multidimensional conceptual framework of servingness at HSIs, which they developed from their systematic content analysis of the HSI scholarship.Specifically, they analyzed how researchers conceptualized serving Latinx students at HSIs (beyond enrollment) and found that they understood servingness in four main ways: (1) outcomes (i.e., academic and non-academic outcomes); (2) experiences (i.e., student and non-student experiences); (3) internal organizational dimensions (i.e., leadership and decisionmaking; culturally relevant pedagogy and curriculum; and culturally relevant programs); and (4) external influences (e.g., public policy).Most commonly, however, they found that researchers conceptualized servingness based on student-level outcomes-their academic outcomes (e.g., graduation rates) and non-academic outcomes (e.g., students' academic self-concept).Based on these findings, Garcia et al. (2019) proposed a multidimensional framework for servingness, which includes indicators of serving and structures of serving.The former refers to measurable individual dimensions, including outcomes and experiences, which communicate if or how well HSIs serve their campus communities.The latter (i.e., structures of serving) includes various organizational dimensions, such as an HSI's compositional diversity, curriculum and pedagogy, programs for minoritized students, and incentive structures.Importantly, these organizational structures affect students, faculty, and staff members' experiences and outcomes.Given most HSIs "minimal history or commitment to serv [ing] Latinx/e students" (Garcia & Cuellar, 2023, p. 2), HSI scholars stress the need for these institutions to transform their organizational structures by "centering Latinx/e ways of knowing and being" (Garcia & Cuellar, 2023, p. 2) to advance servingness.
Altogether, Garcia et al. (2019) offer a comprehensive way of understanding servingness, clarifying what it means or could mean to actually serve Latinx students at HSIs.Among suggestions, for instance, they contend that researchers should take an asset-based/anti-deficit approach when framing studies about HSIs.In line with their recommendations, we apply this framework to extend the research on servingness, specifically to better understand how HSIs serve Latinx and other racially/ethnically minoritized students in STEM.

Methods
To better understand STEM servingness at HSIs, we conducted a systematic review of the literature on STEM undergraduate education at HSIs.Unlike traditional literature reviews that include studies "in an idiosyncratic fashion" (Chong et al., 2022, p. 2), systematic reviews employ a scientific approach to identifying, analyzing, and synthesizing existing literature.Researchers methodically establish search parameters, apply inclusion/exclusion criteria, and then synthesize research on a given topic following a repeatable method (Borrego et al., 2014).In this way, we synthesized what the STEM education literature shows regarding the contributions of HSIs to Latinx and other racially/ethnically minoritized undergraduate students' experiences and outcomes in STEM fields.

Databases and search terms
To begin, we consulted with a university librarian for recommendations on databases, given her expertise in running such queries.Heeding her insights and our informed understanding of the main databases for educational research, including literature on STEM education, as well as modeling other content analyses' data collection processes (e.g., Borrego et al., 2014;Marin & Aguilar-Smith, 2023;McClure et al., 2022), we ultimately landed on the following databases: Education Source, Academic Search Complete, Professional Development Collection, and ERIC.We first used the search terms "Science, Technology, Engineering, Mathematics4 " and "Hispanic Serving" or "Hispanic-Serving" in full text.The search yielded 1,326 articles, most of which were not relevant to STEM undergraduate education at HSIs based on their abstracts.Thus, we used the same terms but narrowed the search for them to publications' titles or abstracts.When we re-ran the query, we also attempted to capture research potentially missed in our initial search given nomenclature issues, adding search terms for STEM subdisciplines, particularly (though not exclusively) ones without science or mathematics as standalone terms.Specifically, we added the following terms: astronomy, biology, biological science, bioscience, bioengineering, chemistry, chemical science, computer science, geoscience, physics, physical science, statistics, and statistical science.The first round of this narrowed search produced 244 publications total, which we transferred from EBSCOhost to RefWorks, a reference management software package.Then, we saved the data to Microsoft Excel for data cleaning.

Selection and screening process Inclusion criteria
In line with the principles of a systematic review, we established inclusion criteria based on our research questions.Specifically, we formulated the following five inclusion criteria (IC): published between 2011 and 2022 (IC 1); written in English (IC 2); contained empirical data published in a peer-reviewed journal or conference proceeding5 (IC 3); focused on undergraduate students or on undergraduate education (IC 4); and addressed STEM disciplines directly (IC 5).We examined the abstracts and methods sections for IC 3-5 and subsequently reviewed the full text to ensure the publications met our inclusion criteria.
To reflect the current state of research, we focused on the past decade of publications (IC 1)-a time of significant growth in both these areas of study: STEM education and HSIs (Li et al., 2020;Marin & Aguilar-Smith, 2023).In particular, given this study's specific focus on STEM education at HSIs, we concentrated on this period because slightly over 80% of all HSI peer-reviewed scholarship came out after 2008 (Marin & Aguilar-Smith, 2023).We also restricted our analysis to publications written in English (IC 2) due to the highly specific context of HSIs within the United States.To put it another way, since HSIs are a U.S. political construct and English is the United States' de facto national language, we reasoned that limiting our search in this way would filter out some unrelated work.
Separately, while systematic reviews often encompass diverse types of work (Borrego et al., 2014), we excluded purely theoretical or historically grounded articles and any non-peer-reviewed sources (e.g., book chapters, dissertations, and reports) per IC 3. We constrained our analysis to exclude such work because peer-reviewed publications are the primary means of disseminating research on STEM education and represent the type of work generally valued within the academy.As such, what these publications say about STEM education and racially/ethnically minoritized STEM students, as well as what they suggest about servingness, holds considerable weight and, thus, aligns with this study's purpose-to attend to STEM education in conversations about servingness at HSIs.
As far as IC 4, we focused on articles related to STEM undergraduate education because this part of a student's academic journey is integral to diversifying the STEM workforce and STEM graduate education.Also, while a segment of HSIs offers graduate/professional degrees, currently more HSIs primarily attend to undergraduate education.Finally, given this systematic review's focus, we excluded articles not addressing STEM or STEM subdisciplines (IC 5).After applying these inclusion/ exclusion criteria, the final dataset included 128 articles.Figure 1-a PRISMA flow diagram-illustrates our search and screening process.

General characteristics of the qualifying articles
To contextualize the findings of this systematic review, we present general characteristics of the 128 qualifying articles.Specifically, we report the following article-level details: year of publication; STEM or sub-STEM disciplines studied; the number of institution(s) studied (i.e., single institutional or multi-institutional study); types of institutions studied; participant population; funding status; and publication outlet.
As shown in Table 1, most of the articles centered on students, though a small proportion focused exclusively on faculty or administrators.In terms of student demographics, about half of the articles explored the experiences of Latinx students alongside other racial/ethnic groups, including whites and Asians.About a third of the articles specifically concentrated on Latinx students, including five publications exclusively studying Latina women.Additionally, about an eighth of the articles broadly focused on racially/ethnically minoritized students, often referring to them as underrepresented minorities (URM) or BIPOC students.
Concerning institutional characteristics, two-thirds of the articles were single institutional studies while a third employed a multi-institutional study approach.Additionally, most articles-a whopping 81%-concentrated on 4-year institutions.Finally, as indicated in Table 1, among the selected articles, half addressed STEM majors in general, while the other half delved into specific STEM subdisciplines such as biology, engineering, or mathematics.
Table 2 goes further still, providing the abridged citation for all the qualifying articles by field.Additionally, in an effort to share more details about the included articles, when we reference any of the qualifying articles in this paper, we mark them with an asterisk (*) in the reference list for readers to easily locate and access.Articles that we did not specifically discuss as examples in this paper are listed in the Appendix.
As far as publication outlets, half of the qualifying articles were published in STEM education journals.However, other scholarly outlets also published this work, including general education journals, Latinx/Hispanic education specific journals, and higher education journals, among others.Table 3 lists the journals where this scholarship has been published to date.

Positionalities
Before delving into our analytic process and approach, it bears first noting our positionalities.Like all scholars, our unique subject positions inform our individual research pursuits and collective thinking relevant to this project (Martínez-Alemán et al., 2015).Specifically, the first two authors of this study are women of color scholars and faculty colleagues who share critical paradigmatic views, despite their generally different methodological approaches and focal areas of inquiry.In particular, the lead author typically employs critical quantitative methods to study issues in STEM, ultimately seeking to diversify and enhance gender and racial/ethnic equity in these areas.Meanwhile, identifying as a Latina woman, the second author generally relies on critical qualitative methods to examine public and institutional policies towards advancing equity and justice, particularly among HSIsinstitutions that overwhelmingly serve her beloved community.The remaining four research team members and authors are doctoral students in the higher education program in which the first two serve as core faculty and higher education professionals with experience supporting minoritized students.With one exception, all of them likewise identify as part of a racially/ethnically minoritized community.Ultimately, despite our diverse social identities and professional roles, we all are cisgendered women who critically study higher education.More importantly, we are collectively invested in advancing equity and social justice within higher education, especially for minoritized students, including Latinx STEM students.Also, of particular relevance to this study, all of us work and/or study at an HSI.Thus, our professional/academic context informs our understanding of and beliefs about servingness at HSIs.Altogether, given our scholarly commitments and current professional/academic context, we are keen on elevating ways to advance STEM servingness at HSIs.

Analytic process and approach
As far as our analytic approach, our research team, which consisted of two faculty members and four doctoral students, systematically followed a protocol-driven process to analyze the qualifying articles in line with our guiding theoretical framework.Specifically, the doctoral student team members initially read the full text of the qualifying studies, summarized each article, and deductively coded the articles in line with the seven components of Garcia et al. (2019) servingness framework.Although they took the lead on this part of the analysis, they did so in close consultation with the lead author and primary investigator of this project.At the same time, while supporting and participating in this deductive phase of analysis as well as regularly checking research SEARCH 244 articles retrieved from Education Source, Academic Search Complete, Professional Development Collection, and ERIC (via EBSCOhost) using the following terms: Hispanic Serving; Hispanic-Serving; Science, Technology, Engineering, Mathematics; sub-STEM disciplines (chemistry, chemical science, biology, biological science, bioscience, bioengineering, physics, physical science, geoscience, astronomy, statistics, and statistical science)

SCREENING
Screened 244 articles' abstracts and methods using inclusion criteria  team members' coding for accuracy, the lead author also independently read and inductively analyzed the articles.Speaking of trustworthiness measures, when there were discrepancies in the coding, we discussed them until everyone agreed on the final coding of each article.
To clarify our analytic process further, we break down and discuss our approach by our two guiding research questions.To answer our first research question (i.e., What components of servingness has STEM education research (under)studied?),we employed a deductive coding approach using Garcia et al. 's (2019) servingness framework (Miles & Huberman, 1994).Specifically, we used this framework's seven components to deductively code each of the qualifying articles: (1) academic outcomes, (2) non-academic outcomes, (3) student experiences, (4) non-student experiences, (4) leadership decision-making, (5) culturally relevant curriculum or pedagogy, (6) culturally relevant practices, and (7) external influences.Furthermore, Garcia et al. (2019) broke these components down, noting examples of each.For instance, academic outcomes include graduation, persistence, course enrollment/completion, and transfer rates.
We used these examples when coding the data.However, we also incorporated additional examples of these seven components that pertain more specifically to STEM.For instance, some of the qualifying articles addressed STEM identity development or STEM self-efficacy-examples not explicitly listed in Garcia et al. 's (2019) model.Accordingly, following their lead, we coded students' academic major-specific psychological attributes (e.g., STEM identity salience or STEM self-efficacy) as nonacademic outcomes.
A few additional points of clarification regarding our deductive analysis process, when we coded student experiences, we first coded whether the article addressed students' classroom or outside-of-classroom (i.e., cocurricular) experiences.Afterward, as we discuss later, we coded whether the experience was validating or racialized.Additionally, a few recent articles in our dataset addressed the effects of COVID-19 on teaching and learning, specifically in STEM classrooms at HSIs.In such cases, we coded COVID-19 as an external influence.Finally, regarding this phase of coding, it bears noting that the articles often addressed multiple components of the servingness framework, meaning that codes are not mutually exclusive.For instance, if an article discussed academic and non-academic outcomes, we applied both codes.Similarly, if an article examined various types of academic outcomes, we took note and coded each unique type of outcome.
Once we completed our deductive coding process, we counted the number of articles that addressed each component and subcomponent of the servingness framework to answer our first research question.In taking this approach, we did not delve into themes within the qualifying articles.Rather, we sought to see-and quantifypatterns in terms of the parts of servingness (under) studied in this set of STEM education literature.
To answer the second research question (i.e., What are gaps in the research on STEM education at HSIs relevant to serving Latinx STEM undergraduates?), we inductively analyzed the 128 articles.As previewed, the lead author largely headed this process, carefully reading the articles and reflecting on them holistically.Specifically, in considering the articles altogether, she wrestled with the following kinds of analytic questions: Where, if at all, does the study address the HSI context-in  third author, who likewise read most of the articles, to talk through and process these analytic questions.After reflecting on these questions and their answers, the lead author worked closely with the second author, an expert in HSIs, to make meaning of and thematically categorize the shortcomings of this literature.In the end, building on our deep knowledge of STEM education and HSIs and guiding theoretical framework, together, we developed themes to describe the gaps within this body of research on STEM education.As presented in the findings, we described the gaps as: ghosting the HSI context; (b) ghosting Latinx culture; and (c) ghosting people and places.

Findings
Guided by our research questions, our systematic review of the literature on STEM education at HSIs revealed the components of servingness studied over the last decade (RQ1) and several gaps within this line of research (RQ2).We begin by presenting the findings of our first research question in line with the components within Garcia et al. 's (2019) servingness framework.Moving on to RQ2, we present the thematic gaps within the selected literature.

Student outcomes
Among the qualifying articles, nearly two-thirds focused on student outcomes.Although codes are not mutually exclusive, the number of articles on academic outcomes  (n = 99) is still nearly twice as high as the number of articles on non-academic outcomes (n = 53; see Table 5).

Academic outcomes
Academic outcomes encompass three sub-categories: student success outcomes, student course performance outcomes, and student learning outcomes (Garcia et al., 2019).Within the subset of articles, approximately half focused on traditional indicators of student success (n = 51), particularly persistence (e.g., Ashford-Hanserd et al., 2020;Contreras Aguirre & Banda, 2019;Mein et al., 2023).On balance, this literature's emphasis on persistence makes sense given the high attrition rates in STEM majors, particularly among BIPOC students (NSF, 2017).
Additionally, approximately a quarter of the qualifying articles focusing on academic outcomes examined student STEM course-level performance indicators (n = 22), including pass/fail rates and students' course grades (e.g., Becker & Cox, 2022;Moreno et al., 2021).Given STEM's notorious weed-out culture and its pronounced negative effect on BIPOC students and women students (McGee, 2016(McGee, , 2020;;Ro et al., 2021), close attention to these course-level indicators tracks, as scholars work to explain and remedy this chronic issue within STEM education.Finally, a quarter of the academic outcome-focused articles addressed student learning outcomes (n = 26).More specifically, many of these studies addressed STEM-relevant learning outcomes, such as students' research skills, design skills, critical and creative thinking, systems thinking, and scientific and quantitative Table 3 Journals identified in the final dataset (N = 128) reasoning.For example, Stanfield et al. (2022) examined systems thinking among students engaged in undergraduate research at an HSI, and Gomez and Svihla (2019) explored research skills among community college students engaged in a collaborative research program with a 4-year HSI.These two studies also examined the students' science identity and other non-academic outcomes associated with participation in undergraduate research.Other studies examined soft skills important for the workforce, including STEM students' teamwork, communication, writing, and project management skills (e.g., Hubka et al., 2019;Jin et al., 2019).For example, Shah et al. (2022) investigated communication skills, focusing on how students translate scientific language and research for community members.Ultimately, attention to these learning outcomes and soft skills is reasonable, considering ABET's accreditation criteria. 6At the same time, this body of STEM education research overlooked other key learning outcomes, such as contextual competence or cultural understanding and interdisciplinary learning.These outcomes are particularly critical in the context of STEM education, as the inclusion of contextual/cultural factors or interdisciplinary team learning in curricula has been shown to be effective in attracting and retaining women and BIPOC students in engineering (Lattuca et al., 2017;Ro et al., 2015).

Student and non-student experiences
Of the 128 total qualifying articles, over half (n = 76) focused on student experiences, while only 10 explored non-student experiences.

Student experiences
As shown in Table 6, out of the 76 articles on student experiences, 28 examined curricular and pedagogical experiences of gatekeeping (e.g., introductory math or physics) or capstone courses (e.g., Bailey, 2019;Hum et al., 2015).Furthermore, within this set of work, 30 studied certain co-curricular programs such as undergraduate research and mentoring programs (e.g., Garibay, 2022;Ing et al., 2021;Marino et al., 2020).Additionally, 18 articles discussed STEM students' broader college experiences at their respective HSIs (e.g., Flores et al., 2020;Kendall et al., 2018;Preuss et al., 2021).Within these same sets of studies, scholars identified validating and racialized experiences that Latinx and other BIPOC STEM students encountered (see Table 7).According to Garcia et al. (2019), a core component of servingness at HSIs is students' college experiences, specifically whether it is validating or racializing.Validating experiences include, for example, Latinx students seeing themselves reflected in their faculty and peers.Meanwhile, racialized experiences include, for example, Latinx students encountering discrimination, harassment, or racial microaggressions.
Curricular and co-curricular experiences Given Garcia et al. organizational-level analysis, they did not focus on classroom-level experiences.However, within our systemic review, we found that a third of articles on student experiences focused specifically on experiences in classroom settings (n = 28).Furthermore, nearly all of these articles focused on students' curricular and/or pedagogical experiences.More specifically, these studies addressed students' experience with active learning pedagogies (e.g., Davis, 2017), online learning tools (e.g., Altowaiji et al., 2021;Broussard et al., 2021;Thacker et al., 2022), and various assessment methods (e.g., Henderson et al., 2019).
Of the 76 articles on student experiences, 30 addressed specific co-curricular programs designed to support BIPOC STEM students.Most of this scholarship centered on the impact of participating in undergraduate research on student outcomes.However, as a standout case, Haeger et al. (2021) addressed both students' access to these programs as well as their effects.In terms of the former, Haeger et al. identified structural and cultural barriers to participation in undergraduate research, including inadequate funding and faculty members' selection process of team members.Given that STEM programs often offer undergraduate research opportunities (Carter et al., 2016), it is understandable that we found 18 articles focused on the impact of such programs on minoritized students at HSIs (e.g., Shuster et al., 2019;Yedinak et al., 2018).In contrast, while internships are a common high-impact practice within STEM education (Kilgo et al., 2015), only three articles mainly addressed STEM students' internship experiences and outcomes (i.e., Enriquez et al., 2015;Langhoff et al., 2017;Sansone et al., 2019).Additionally, collectively, only nine studies addressed student involvement in support services and programs for BIPOC students (e.g., Bouniaev et al., 2014;Casey et al., 2019;Jin et al., 2019;Khasawneh et al., 2014).
Ultimately, while these articles addressed important college experiences, they largely failed to consider the unique context of HSIs or the cultural backgrounds of Latinx or other BIPOC students at these institutions.Given that many studies acknowledged within their limitations the lack of generalizability of their work due to focusing on a single institution and limited information about the institutions and students under study, it is unclear how the findings of these studies apply to other HSIs.Validating and racialized experiences Out of the 76 articles related to student experiences, about half (n = 41) highlighted validating experiences with peers and faculty in their STEM majors and professionals in STEM fields (see Table 7).Among this subset of articles, 21 discussed STEM student interactions with STEM faculty, overwhelmingly highlighting the positives of these interactions (e.g., Bailey, 2019;Biswas et al., 2015;Rodríguez Amaya et al., 2018).Additionally, nine articles addressed the role and impact of peer mentorship or supports for Latinx STEM students (e.g., Cruz et al., 2021;Gates et al., 2011).Finally, 11 articles addressed Latinx student experiences of cultural validation inside and outside of the classroom, such as their course's relevance to the Latinx community (e.g., Mein et al., 2023;Monarrez et al., 2024;Preuss et al., 2021).Considering that these studies exclusively focused on students' perspectives, the absence of faculty-side data or narratives on these experiences (as well as their own) is notable (as previously mentioned, only 9 out of 128 articles delved into non-student experiences; refer to Table 4).While recognizing faculty members' work to support and mentor students, the extent to which HSIs have set policies and practices to serve STEM students, particularly minoritized STEM students, as opposed to relying on the good-faith efforts of individual faculty mentors, remains unclear.
Separately, despite much of the literature documenting minoritized students' negative experiences in STEM (Ro et al., 2021;Seymour & Hewitt, 1997), we surprisingly found that only nine articles primarily addressed Latinx STEM students' racialized experiences and the adverse influence of these negative experiences (e.g., Ashford-Hanserd et al., 2020;Camacho & Lord, 2011;Martinez Ortiz & Sriraman, 2015).These studies highlighted Latinx STEM students' experience with (co)overt forms of discrimination, STEM's exclusionary culture, racial microaggressions, and acculturative stress.Furthermore, and importantly considering servingness, it is worth noting that 26 of the 76 articles on curricular experiences did not explore whether minoritized STEM students experienced (in)validation based on their cultural backgrounds or identities.

Non-student experiences
In addition to student experiences, Garcia et al. (2019) incorporated the experiences of faculty, administrators, and staff as an integral component of servingness.Compared to the 76 unique articles on student experiences, only 9 unique articles included the perspectives and experiences of faculty, staff, and administrators (see Table 4).These articles concentrated on how STEM faculty members try to enhance their understanding of Latinx students toward improving their teaching and mentorship and, ultimately, Latinx STEM student success.For example, Biswas et al. (2022) documented how 40 STEM faculty members at an HSI worked to reform their approach to teaching and learning, adopting evidence-based, student-centered practices to ultimately better support their students.

Internal organizational dimensions
As a reminder, Garcia et al. (2019) categorized internal organizational dimensions as culturally relevant pedagogy and curriculum; culturally relevant programs; and leadership and decision-making.Out of 128 total unique qualifying articles, we identified only 14 articles on culturally relevant curriculum and pedagogy in STEM (see Table 4).Among these 14 articles, only three covered curricular content that specifically centered on Latinx students' cultural interests (i.e., Bhattacharya et al., 2022;Hammersley et al., 2012;Race et al., 2021).Additionally, 11 studies focused on pedagogies that embrace an assetbased understanding of Latinx students (e.g., Cavazos et al., 2022;Gomez et al., 2018;Zeng & Zeng, 2021).
In comparison to culturally relevant pedagogy and curriculum, we found more focus within this literature on culturally relevant programs-programs seeking to support Latinx and BIPOC students by recognizing their cultural and racial values.We identified 30 articles about outreach, retention, and co-curricular programs for BIPOC students largely funded by external agencies (e.g., Biswas et al., 2015;Camacho et al., 2014;Casey et al., 2019).These programs are designed to provide Latinx and other BIPOC students with opportunities for mentoring, tutoring, undergraduate research, internships, and faculty interaction.For example, Enriquez et al. (2014) explored paid internships as a way to address many Latinx STEM students' financial constraints.Additionally, scholars studied same-race or same-gender peer mentors (e.g., Cruz et al., 2021) and teaching assistants (e.g., Lee & Ing, 2020) to foster a greater sense of belonging for BIPOC STEM students (e.g., Burn et al., 2019;Shultz et al., 2022).
Lastly, as far as internal organizational factors relevant to servingness, Garcia et al. (2019) highlighted HSIs' leadership and decision-making.Out of the 128 unique qualifying articles, we only found eight articles that primarily addressed leadership and decision-making (see Table 4).These studies focused on the campus leadership's development of certain STEM initiatives, changes in institutional and systemic practices, and the formulation of strategies to recruit and retain Latinx students at HSIs.For example, Núñez et al. (2021) and Gates et al. (2011) exemplified how HSI leaders have developed new strategies and programs as part of their involvement in CAHSI, a network involving over 60 HSIs and other stakeholders dedicated to advancing Hispanic achievement in computing.Garcia et al. (2019) conceptualized external factors as those that "have affected the development and growth of HSIs, including decisions and policies of the local, state, and federal government; factors beyond the control of the institution" (p.758).In our review, most articles briefly mentioned state and federal legislation as part of HSIs' broad background, but these external factors were not the primary focus of these STEM education studies.And so, in line with Garcia et al. (2019), we found a limited number of publications addressing external factors (n = 9; see Table 4).Notably, most of these articles focused on the impact of the COVID-19 pandemic on STEM students' learning experiences and outcomes without particularly considering the uniqueness of the HSI context (e.g., Broussard et al., 2021;Contreras, 2017;Morales-Cruz et al., 2021).

Thematic gaps within STEM education research at Hispanic-serving institutions
As prefaced, in response to our second research question (i.e., What are the gaps in the research on STEM education at HSIs relevant to serving Latinx STEM undergraduates?), we identified three key thematic gaps in the literature on STEM undergraduate education at HSIs, which we describe as (a) ghosting the HSI context; (b) ghosting Latinx culture; and (c) ghosting people and places.

Ghosting the HSI context
Similar to what other scholars have found in their systematic reviews of the broader HSI scholarship (e.g., Garcia et al., 2019;Marin & Aguilar-Smith, 2023), we found that many STEM education studies, both quantitative and qualitative alike, relied on HSIs as a convenient research site without considering HSIs' unique context.Specifically, although all the articles mentioned HSIs either in their abstracts or keywords, only a third (n = 44 articles) mentioned the relevance or contribution of HSIs in any depth somewhere in their methods, findings, discussions, and/or implications sections.That is, only a third of the articles drew from HSI scholarship, discussed their findings considering the HSI context, or offered implications expressly for HSIs and their campus communities.Instead, the bulk of the articles simply mentioned HSIs as the study site, mainly within their methods section.Often, in this particular set of studies, researchers appeared to select an HSI as the study site because of pure convenience (i.e., one of the researchers worked at the institution in question) or the institution's convenient student demographics.In terms of the latter, these researchers seemingly selected an HSI given their interest in studying some phenomena related to Latinx STEM students (e.g., Casey et al., 2019;Fernández et al., 2021;Flores et al., 2014).
Although this body of STEM education research largely ghosted the HSI context, quantitative studies ghosted this context more than qualitative studies.Specifically, quantitative studies relying on classroom-or course-level evaluation data tended to underplay the role of context.Consequently, they may fall short in providing implications keenly relevant to HSIs.Furthermore, we found that articles published in education or Hispanic-specific journals tended to better consider the HSI context.For instance, these studies generally devoted ample space to HSIs' unique characteristics, histories, and student profiles (e.g., Bensimon et al., 2019;Herrera & Kovats Sánchez, 2022;Sansone et al., 2019).
Within this theme of ghosting context, we also observed that many of the articles on STEM student outcomes and classroom experiences underemphasized or altogether overlooked the students' connection to the broader HSI campus or community.Additionally, multiinstitutional studies on student outcomes often failed to contextualize the potential differences in outcomes between Latinxs at HSIs and their counterparts at PWIs.Put differently, they dismissed the explanatory possibilities of these differences tied to HSIs' unique conditions.Similarly, single institutional studies on curricular or pedagogical practices in our dataset were largely silent on how these practices do or could effectively work in the context of HSIs.Relatedly, while some articles explained that HSIs' compositional diversity, specifically their sizeable Latinx student populations, may positively connect to student outcomes (e.g., Flores et al., 2014;Navarro et al., 2014), they generally did not delve into the dynamics of learning and development among Latinx students within the HSI context.
Ultimately, while choosing HSIs as a study site as a means of addressing a pressing scholarly question is not inherently problematic, the lack of deliberate consideration of the context of HSIs is an issue-one jeopardizing STEM servingness.Additionally, this overriding decontextualized approach in this strand of STEM education research erroneously suggests that HSIs are unidimensional and monolithic in nature.Essentially, much of this research neglects the significant diversity within this population of over 500 colleges and universities and, moreover, the importance of these differences to the very questions undergirding their studies.Akin to what McClure (2018) noted about the research on public regional universities, the net effect is that "this tendency towards aggregation reinforces the trope that these institutions are identity less" (p.120).In doing so, the work under-realizes its potential to enhance STEM education and serve STEM students.

Ghosting Latinx culture
While the majority of articles focused on STEM student experiences, most articles did not center on the sociocultural aspects of Latinx students or communities.For example, 14 articles examined traditional gatekeeper courses in STEM education (e.g., entry-level mathematics and science courses) given racially/ethnically minoritized students' overall underperformance in these critical courses (e.g., An et al., 2022;Convertino et al., 2022;Vemu et al., 2022).However, these studies generally did not address whether or how the content or pedagogical approaches within these gatekeeper courses connected to Latinx culture or communities.Indeed, we only identified two articles that considered Latinx students' cultural backgrounds in relation to the curricular content of gatekeeper courses (i.e., Bhattacharya et al., 2022;Hammersley et al., 2012).For example, Bhattacharya et al. ( 2022) implemented decolonial project-based learning in precalculus classes (i.e., typically a gatekeeping course, adversely affecting aspiring BIPOC STEM students), incorporating social justice related and culturally connected topics into these courses.
Nevertheless, despite extensive research underscoring the importance of connecting the classroom to students' cultures, particularly for Latinx college students (Garcia & Okhidoi, 2015;Pappamihiel & Moreno, 2011), the articles we reviewed, including ones on curricula and pedagogy, largely glossed over students' cultural backgrounds.In doing so, they rendered it unclear if, how, or why these curricular and/or pedagogical practices work for Latinx STEM students at HSIs.
Similarly, articles focused on co-curricular experiences paid limited attention to Latinx students' culture.Specifically, only a few studies considered whether co-curricular opportunities embed Latinx culture or whether STEM faculty or peers understand Latinx STEM students' cultural and social identities.For example, Haegar et al. ( 2021) identified programmatic, cultural, and environmental barriers to participation in undergraduate research opportunities.Among barriers, they found that research programs' largely individualistic cultures deter many Latinx students.Similarly, Rodríguez Amaya et al. (2018) reported that many Latinx students do not participate in undergraduate research because they think research is done alone, without collaboration among peers and faculty mentors.In effect, these studies point to the perceived cultural incongruence between co-curricular STEM programs and Latinx students, who tend to be highly group-oriented and value collectivism (Núñez et al., 2021;Preuss et al., 2021;Sáenz & Ponjuán, 2011).Ultimately, without a clear understanding of Latinx culture, including its strong collectivist orientation, STEM departments risk perpetually under-enrolling and underserving this long-underrepresented community of students in STEM.
Lastly, although many studies noted the importance of student-faculty interactions for STEM student success, particularly how STEM professors' interest and support positively impact minoritized students, they rarely mentioned if or how understanding the cultural and social aspects relevant to Latinx and BIPOC students informs these critical relationships.Furthermore, these articles did not substantially explore the racial or gender dynamics within these student-faculty interactions, although extensive research underlines the importance of these considerations in student success, particularly for BIPOC college students (Contreras Aguirre et al., 2020, 2021;Gonzalez et al., 2021;Herrera et al., 2022).In fact, based on our review, Contreras Aguirre and Banda (2019) were one of the few scholars to talk about the need for faculty mentors with shared racial and/or gender identities to support Latina women STEM students.However, Brown et al. (2020) did suggest that faculty receive training on how to ethically interact with mentees of different cultural backgrounds.

Ghosting people and places
The STEM education research also ghosted groups of people and places.While many of the articles discussed Latinx and other BIPOC students, the literature underexamined certain student groups.Specifically, only five articles addressed Latina women college students specifically and highlighted their experiences with racialized sexism in STEM environments (e.g., Acevedo et al., 2021;Contreras Aguirre et al., 2020;Frederick et al., 2020).Given the overriding underrepresentation of women, particularly women of color, in STEM fields (Ong et al., 2011(Ong et al., , 2018;;Ro et al., 2021), the limited attention to Latina women students within STEM education research at HSIs is concerning.In addition, the research on STEM education neglects Latino men.Indeed, none of the qualifying articles focused on Latino men STEM students at HSIs.
Furthermore, it is important to note that HSIs enroll not only Latinx students, but also students from other racially/ethnically minoritized groups (Excelencia in Education, 2023b).However, most studies in our review primarily focused on Latinx students, with only two exceptions.Specifically, Kodaseet and Varma (2012) explored the experiences of American Indian students in computer science programs at both HSIs and Tribal Colleges and Universities, and Williams Pichon (2016) examined Black STEM students' sense of belonging compared to those in non-STEM majors at an HSI.
In addition to inattention to particular student groups, the experiences and perspectives of other members of the campus community have been relatively understudied with the STEM education research.Again, only 10 articles in our dataset addressed STEM faculty members' teaching and mentoring experiences.Also, there has been limited attention to administrators' and staff members' leadership and decision-making.Similarly, the roles and contributions of professionals in the STEM industry are not well addressed.In fact, we found that only Bensimon et al. (2019) addressed the roles of STEM faculty, administrators, and staff in promoting the persistence of Latinx STEM students.
Finally, this literature ghosted places, too.Specifically, only 24 articles included Hispanic-serving community colleges (HSCCs) in their analyses.Half of those explicitly focused on HSCCs (e.g., Langhoff et al., 2017;Vemu et al., 2022), and the other half included data from both 2-year and 4-year HSIs (e.g., Biswas et al., 2015;Froyd et al., 2015).Some studies addressed collaboration between faculty at HSCCs and 4-year HSIs to offer research opportunities to community college students (e.g., Enriquez et al., 2015;Loeser et al., 2021;Rentsch et al., 2014).In the end, though, there remains a significant gap in the research regarding Latinx and other BIPOC STEM students' outcomes and experiences at HSCCs and their transitions from 2-year to 4-year institutions.Consequently, the contributions of HSCCs and the systemic barriers faced by their students and the broader STEM campus community remain largely unknown.

Discussion
Through our systematic review of the STEM education research at HSIs, we addressed which components of servingness have been (under)studied (RQ1) and identified limitations and gaps within this literature (RQ2).In attending to this disciplinary context, we extend Garcia et al. 's (2019) servingness framework, which conceptualized servingness at the institutional level; we return to this point before addressing implications for research, policy, and practice.
As a reminder, in terms of the first research question, we found that most of the studies unsurprisingly focused on students' outcomes and experiences-an overriding concern within STEM education research more broadly (Li et al., 2020;NSF, 2022).Indeed, scholars studying Latinx STEM students' persistence and degree completion are poised to improve STEM programs and the workforce, helping address, for example, the underrepresentation of Latinx engineers and scientists.These studies also extensively examined how classroom learning, co-curricular programs, and support systems help or hinder Latinx STEM students succeed.In doing so, these studies support STEM faculty and administrators in program development and implementation.Despite these notable contributions, the studies often overlooked the HSI context generally or the institution's specific organizational context, which raises questions about the effect(iveness) of the various educational interventions studied within this literature.
Beyond the numerous student-focused articles, relatively few STEM education studies explored HSIs' internal organizational dimensions, including the composition and role of faculty and staff and external factors in serving STEM students.Although this finding aligns with Garcia et al. 's (2019) systemic review of the HSI scholarship, the scarcity of current literature in this area is concerning given preliminary research showing that STEM faculty and administrators are change-makers.Specifically, they help shape the vision, culture, policies, and practices of STEM programs at HSIs (Bensimon et al., 2019;Kang et al., 2022) as well as serve as mentors for Latinx and other BIPOC students (Biswas et al., 2015;Bouniaev et al., 2014;Casey et al., 2019;Contreras Aguirre & Banda, 2019;Contreras Aguirre et al., 2020;Strong et al., 2019).Also, external factors, including the political climate or relationship with funding agencies, undoubtedly influence faculty and administrators' work, students' learning conditions, and ultimately an institution's STEM servingness.Thus, failing to account for these factors leaves unclear potentially key components in advancing STEM education, including how STEM faculty and administrators enhance STEM education and foster diversity in the STEM workforce.
Moving to our second research question, we found that in many cases, STEM education studies ghosted or overlooked the HSI context, Latinx culture, and key people and places.In line with findings from previous systematic reviews of HSI scholarship (Garcia et al., 2019;Marin & Aguilar-Smith, 2023), we identified that many STEM education studies used HSIs as a way to access Latinx students and neglected HSIs' distinct context.Yet, HSI scholars have repeatedly stressed the importance of considering HSIs' unique organizational context and realities to better understand and serve these institutions and their campus communities (Garcia, 2015;Núñez, 2017).For example, while HSIs generally support students with greater needs, they typically have fewer resources than PWIs (Nellum & Valle, 2015;Ortega et al., 2015); these resource conditions especially impact STEM.STEM programs are resource-intensive enterprises, requiring lab facilities, top-notch technology, and extensive equipment (NASEM, 2019;NSF, 2021).Hence, considering an HSI's financial circumstances is crucial when considering possibilities for STEM education within this context.
Additionally, failing to consider the unique context of HSIs can lead to a misunderstanding of Latinx students' needs.For instance, while students at HSIs often find themselves in a racially/ethnically diverse campus environment, research suggests that many students still encounter racialized experiences (e.g., racial microaggressions and anti-Blackness) on campus (Abrica et al., 2020;Cuellar & Johnson-Ahorlu, 2016;Serrano, 2022).These racialized experiences weigh on students' academic lives.Without understanding the complexities of this context, STEM education research might overlook the racism and racialized experiences faced by Latinx students even at HSIs, and this oversight can hinder Latinx STEM students' persistence and success.
In addition to ghosting the context, we also observed a lack of consideration for Latinx culture across the STEM learning environment, including within STEM curricula, pedagogical practices, and co-curricular programs.Yet, budding research has demonstrated the value of integrating Latinx culture into STEM learning for Latinx students' engagement and success (López et al., 2019;Núñez et al., 2021).Ultimately, inattention to Latinx culture in STEM education may alienate and invalidate Latinx students within these environments and, thus, hinder efforts to diversify STEM programs and the STEM workforce.
Finally, our findings revealed additional gaps in this line of research, including a narrow focus on Latinx students only, treatment of Latinx students as a monolithic group, and a lack of research on HSCCs.While HSIs enroll and serve students of varied racial/ethnic backgrounds, most studies focused on Latinx students or studied Latinx students alongside other racial/ethnic groups (see Table 1).Furthermore, Latinx students are a diverse pan-ethic community, including people with varied cultural and national heritages, immigrant status, phenotypical presentations, languages, social class, and gender identities (Excelencia in Education, 2023b; Garcia & Cuellar, 2023).However, these studies overwhelmingly approached Latinx students as a homogenous group, except for the rare article that addressed Latinx STEM students from borderland regions (i.e., Convertino & Mein, 2017) or Latina women STEM students with diverse backgrounds (i.e., Contreras Aguirre, 2023).Separately, the emphasis on 4-year institutions within this literature means there is a lack of understanding of what happens at HSCCs regarding STEM education.This tendency is concerning since millions of Latinx and other BIPOC students start their college careers at community colleges (American Council of Community Colleges, 2024) and since burgeoning research shows that HSCCs open gateways to STEM fields and careers for BIPOC students (Crisp et al., 2009;Herrera & Rodriguez-Operana, 2020;Zamani-Gallaher et al., 2019).
Finally, beyond our findings related to our two research questions, our study extends Garcia et al. 's (2019) multidimensional framework of servingness by attending specifically to STEM education within the context of HSIs.Despite the framework's comprehensiveness and usefulness in helping operationalize servingness, we identified another key area that deserves inclusion in the framework: academic and disciplinary environments.Since the framework was initially developed for university-wide, organizational-level analyses of servingness at HSIs, it did not explicitly account for potential differences between academic disciplines or disciplinary environments.However, there is reason to suspect that these contexts may interface with servingness, as students' interests, needs, learning experiences, and perceptions of campus differ by their academic majors (Holland, 1996;Lent & Brown, 2019).Also, faculty culture, norms, compositional diversity, and policies and practices vary between and within academic disciplines (Lattuca & Stark, 2009).For example, given current gender and racial differences in the faculty composition within STEM fields, Latina women biology majors have greater odds of identifying faculty role models that share their identities than Latina women engineering students.Ultimately, attending to disciplinary differences is important to advance servingness at HSIs and STEM servingness in particular.

Limitations and implications for future research
The limitations and findings of this study present ample opportunities for future research.As a reminder, for this study, we focused on peer-reviewed journal articles and conference proceedings published from 2011 to 2022.Thus, we omitted insightful work on this topic (e.g., books, book chapters, dissertations, and reports), which would provide a fuller understanding of STEM servingness at HSIs.Hence, building off this study, future scholars should run a more expanded search to capture recently published work on this topic.Indeed, with the substantial growth in STEM education and HSI scholarship in recent years (Marin & Aguilar-Smith, 2023), this expanded timeframe would prove useful.Additionally, future scholars should apply less restrictive inclusion criteria, including, for instance, other valuable forms of knowledge, such as edited volumes, monographs, and reports, on this topic in their analyses.
Aside from opportunities for research in light of this study's limitations, our findings also highlight several lines for future inquiry.First, considering the swath of studies in our dataset that seemingly used HSIs to study something else, namely Latinx students, we recommend that researchers clarify their reasons for choosing an HSI(s) as the study site before adding "HSI" to the article's title and keywords.If the primary focus of their studies is not HSIs, then it is rather misleading to foreground HSIs in the title and/or abstract.Even more, it calls into question the intention of this move.Does it reflect the authors' interest in maintaining the appearance of equity and diversity?Meanwhile, if the study is about STEM education at HSIs, researchers ought to engage the growing HSI literature.Moreover, as Núñez (2017) explains, it is paramount that scholars consider the HSI context and what this context may suggest about their study's findings and implications.
Second, considering the inattention to several components of servingness within STEM education research based on our systematic analysis, scholars should incorporate internal organizational dimensions of HSIs or structures for serving in their analyses.Relatedly, identifying relatively few studies on non-student experiences, curriculum and pedagogy, and leadership and decisionmaking within this body of work, it is unclear if/how STEM faculty and administrators at HSIs implement diversity, equity, and inclusion (DEI) initiatives within their programs and departments to diversify STEM.For instance, some colleges may codify these efforts within their mission and vision statements whereas other colleges may carry out this work more covertly, if at all.As a starting point, though, future researchers could examine the mission and vision statements and strategic planning efforts across STEM departments at HSIs.
Another fruitful area of research regarding internal organizational structures includes STEM departments' compositional diversity for both faculty and students.Despite boasting anywhere from 25%-100% Latinx undergraduate student enrollment rates (Excelencia in Education, 2023a), the professoriate at HSIs is far less racially/ethnically diverse (Banda et al., 2017;Contreras, 2017;Vargas et al., 2020).Given extensive research touting the importance of gender and racial/ethnic representation among faculty for student success (Main et al., 2020) and the lack of attention to faculty demographics in STEM education research, at least within the context of HSIs, we encourage future researchers to examine the compositional diversity among the professoriate across STEM and among STEM subdisciplines.Additionally, scholars ought to tease out ways to increase the number of Latinx and other BIPOC in STEM programs-a move that would support the recent efforts by the Alliance of Hispanic-Serving Research Universities to diversify the professoriate and graduate study body.
Third, to advance STEM servingness, external factors, such as public policy and the role of funding agencies, merit greater attention in STEM education research, as these existing studies have largely overlooked these factors based on our analysis.For instance, while recent studies have brought up threats to higher education's mission of diversity in general (e.g., Abrica & Oliver Andrew, 2024;Miller et al., 2023), the STEM education literature has not largely attended to these externalities and their implications for STEM learning environments, particularly for racially/ethnically minoritized STEM faculty and students.Yet, acknowledging ongoing societal issues and contending with the prevailing sociopolitical climates within STEM classrooms and co-curricular programs is poised to strengthen these programs.Research has indicated that Latinx and other BIPOC students are attracted to and more engaged in STEM programs if they are contextually and culturally relevant (Shultz et al., 2022).Along the same lines, recognizing that STEM students and faculty benefit from university-wide DEI programs and support services (Ro et al., 2022;Starr et al., 2022), scholars ought to examine how recent anti-DEI legislation (i.e., an external force) jeopardizes STEM servingness at HSIs.Ultimately, attending to external influences on STEM education can aid STEM faculty and researchers in designing curricula and programs that can better serve Latinx and other BIPOC STEM students.
Fourth, our finding that the STEM education research, to date, ghosts whole groups of people and places reveals prime opportunities for future research.For instance, given our findings and the underrepresentation of women, particularly women of color, in STEM (Acevedo et al., 2021;Contreras Aguirre et al., 2020), it is imperative to conduct more studies on Latina women students' experiences and outcomes in STEM.Our analysis also points to the pressing need to better understand the experience of Latino men in STEM-a student population entirely neglected within the extant STEM education research at HSIs.Even more, future studies should go beyond these gender binaries to ultimately better serve all STEM students.Relatedly, in finding that most of the articles ghosted Latinx culture, our study also underlines the need to deepen the understanding of Latinx STEM students.The Latinx umbrella encompasses a vast array of people with varied cultural and national backgrounds, immigration statuses, racial identities, language proficiencies, etc., which calls for more intersectional analysis of these students and the organizations serving them (Garcia & Cuellar, 2023).
STEM education research also overlooks other minoritized populations (e.g., non-Latinx BIPOC students); this oversight needs to be addressed in order to advance servingness within STEM programs and to diversify STEM.Finally, given the contribution of HSCCs to Latinx education and STEM education (Herrera & Rodriguez-Operana, 2020;Zamani-Gallaher et al., 2019), and, that HSCCs represent the largest single segment of the HSI population (Excelencia in Education, 2023a), more studies are needed to examine servingness, particularly STEM servingness, within this sector.

Implications for policy and practice
As noted, various qualifying articles addressed programs largely funded by external agencies, especially NSF.Indeed, funding agencies like NSF play a pivotal role in supporting STEM education research and expanding Latinx participation in STEM (NSF, 2022).However, in light of our findings, particularly the limited attention to how STEM programs at HSIs may genuinely better serve Latinx STEM students, we encourage funding agencies to prioritize grant proposals that intentionally address how their projects contribute to servingness and, in particular, how their proposals are poised to advance their institution's or their field's ability serve minoritized STEM students.Funding agencies should continue to offer various opportunities for researchers to better understand the Latinx community.For example, the NSF-funded project UNIDOS I Program Network Resource Center for Community Coordination aims to build networks across HSI communities that are "for HSIs by HSIs" (Cardella et al., 2023).Lastly, regarding funding agencies, we suggest that they also offer professional development opportunities for STEM scholars and educators at HSIs to better understand the HSI context and Latinx students, given the decontextualized approach across most of the STEM education scholarship we analyzed.In the long term, such training will hopefully support Latinx students' success in STEM education.
Also, in terms of policy, HSCCs play a vital role in assisting Latinx students in obtaining associate degrees and transferring to 4-year institutions (Herrera & Rodriguez-Operana, 2020;Zamani-Gallaher et al., 2019), but they face pronounced resource and funding challenges (Marin & Aguilar-Smith, 2023).For instance, community colleges often cannot offer internships and research opportunities, although such programs have long been recognized as beneficial for STEM student success and learning (Enriquez et al., 2015;Ro et al., 2015).Furthermore, given that institutional resources aid community college students' transition to 4-year HSIs (Ogilvie & Knight, 2019), we urge policy actors to increase financial support and funding opportunities for HSCCs and their STEM programs.In short, increased funding is necessary to enable these at-promise institutions to expand on the good work they are already doing in providing pathways to STEM degrees and careers for Latinx and other BIPOC students.
As far as practice, we stress again the importance of STEM faculty members, administrators, and staff better understanding Latinx students' cultural backgrounds, as this knowledge can enhance Latinx STEM students' academic and social validation within and beyond the classroom, a key to student success.Echoing our suggestion that funders offer professional development, we similarly encourage HSIs to provide such education to their administrators, faculty, and staff.In particular, we see prime possibilities for such professional development between STEM programs and offices of institutional research, faculty success, and/or teaching and learning at HSIs.In addition to campus-wide efforts, STEM leaders at HSIs may also benefit from actively engaging in national organizations geared to Latinxs in STEM like the Society for the Advancement of Chicanos/Hispanics in Science (known as SACNAS) or invested in promoting STEM education at HSIs like Excelencia in Education, the Hispanic Association of Colleges and Universities, and the Alliance of Hispanic Serving Institution Educators, among others.
Separately, attempts to instill technical knowledge and fundamental skills in STEM courses are frequently done via large lectures that deemphasize the social context and role of social identities (Smith et al., 2016).However, this instructional approach tends to normalize whiteness and reinforce the dominance of white men in these environments to the detriment of women and BIPOC people in these fields and careers (Contreras Aguirre et al., 2020).
To combat this issue and advance STEM servingness, we echo Núñez et al. (2021) who proposed intentionally integrating elements of Latinx culture, particularly confianza, respeto, and familismo, within STEM education at HSIs.Collectively, these campus-wide efforts can help foster a more diverse and inclusive STEM education and workforce.

Conclusion
Through our systematic review of the STEM education research at HSIs, we identified (under)studied components of servingness and gaps within this literature base and extend Garcia et al. 's (2019) multidimensional servingness framework.Even more, finding that STEM scholars have largely neglected HSIs' unique context and Latinx students' culture, we elevate opportunities to advance STEM servingness and, ultimately, improve STEM education and diversify STEM programs and the STEM workforce.In the end, we identify shortcomings with this literature and present opportunities not in a spirit of critique but of possibility for positively transforming STEM education-with your help.

Table 4
Chamely-Wiik et al., 2020;Henderson et al.et al. 's (2019)study.In short, like Garcia et al., we similarly found that the bulk of the qualifying articles concentrated on student outcomes and experiences, though a few studies (n = 9) addressed non-student experiences, particularly those of STEM faculty members at HSIs (e.g.,Chamely-Wiik et al., 2020;Henderson et al.

Table 2
Qualifying articles by field (N = 128)

Table 4
Components of HSI servingness within STEM education research, 2011-2022a An article can address more than one (sub)component of servingness, so the number of articles exceeds 128-the total number of unique articles within the dataset

Table 5
Student academic and non-academic outcomes a An article can address more than one (sub)component of servingness.As such, we counted each sub-category separately, meaning an article may be counted more than once

Table 6
Student curricular, co-curricular, and general college experiences

Table 7
Validated experiences and racialized experiences