Introduction

Higher education has grappled for decades with diversifying the science, technology, engineering, and mathematics (STEM) field pipeline by increasing the rates of underrepresented undergraduate students who enroll in and successfully complete STEM degrees. Federal agencies, foundations, and corporations have invested billions of dollars in STEM education and research to address these disparities, with a particular focus on racially minoritized (RM) students and women (National Science and Technology Council, 2011; Rincón & George-Jackson, 2016). Yet, a 2021 Pew Research Center report shows the persistence of uneven progress in closing these gaps in representation: Black workers make up 9% of the STEM workforce even though they comprise 11% of all employed US adults (Fry et al., 2021). Similarly, Hispanic workers represent 17% of the general workforce but only 8% of the STEM workforce. And although women have in recent years entered STEM fields in great numbers, this progress is concentrated in health-related jobs as opposed to fields like engineering and computing (Fry et al., 2021).

Demographic disparities in college enrollment and completion rates are not specific to STEM disciplines. As of 2022, for example, the overall postsecondary graduation rates of Hispanic, Black, and American Indian students in the USA continue to lag that of Asian and White students (NCES, 2022). However, research has shown bias and discrimination to be particularly prevalent in STEM programs (McGee, 2016). Riegle-Crumb et al.’s (2019) study further revealed a persistent, disproportionate exclusion of Black and Latino/a students in degree attainment that is unique to STEM and not other fields of study. The authors found that STEM is the only field where Black and Latina/o students are significantly more likely than their White counterparts to change majors. Moreover, adding socioeconomic class as a control variable in their statistical models explained these differences for Latina/o students, but no factors—not even high school academic preparation—accounted for disparities in switching majors out of STEM for Black students. Similar trends also applied to the outcome of degree attainment. While social background variables fully explained racial differences in degree attainment rates for non-STEM fields, this was not the case for STEM programs. In other words, STEM programs seem to present unique barriers for students with minoritized identities that are different from inequalities seen across other postsecondary fields of study.

Many studies have thus examined the implementation of “high impact practices” (Kuh, 2008) like learning communities, peer collaboration, student internships, and research experiences designed to increase minoritized students’ enrollment and performance in STEM programs (Chang et al., 2014; Crisp et al., 2009; Espinosa, 2011; Estrada et al., 2016; Graham et al., 2013; Hathaway et al., 2002; Ives et al., 2023; Russell, 2017). Multi-component STEM intervention programs that combine supplemental learning, mentorship, skill building, financial aid, socializing, and a transitional bridge between institutions also show promise in improving the outcomes of minoritized undergraduate STEM students (Palid et al., 2023). Institutional course enrollment patterns can matter as well; promoting the representation of RM students, females, and first-generation college students within STEM courses has been positively associated with higher course grades for all students (Bowman et al., 2022, 2023).

We, however, argue that more attention needs to be paid to other dimensions of equity besides achievement and representation. The National Academies of Sciences, Engineering, and Medicine (NASEM)’s 2022 consensus study report called for four approaches to equity that include the following: “1) increasing opportunity and access to high-quality science and engineering learning and instruction; 2) emphasizing increased achievement, representation, and identification with science and engineering; 3) expanding what constitutes science and engineering; and 4) seeing science and engineering as part of justice movements” (p. 23). Whereas increasing access, achievement, and representation are important tenets of STEM equity, the report underscores how the pursuit of all four approaches in tandem are necessary to achieve not just equity, but also justice within the field of STEM. Yet, there has been little progress or emphasis on the latter two approaches in particular.

To this end, we seek to spur deeper discourse on these two underdiscussed areas of what constitutes STEM and the role that STEM plays in broader social justice through the lens of changes to STEM pedagogy. In spite of strong evidence against the use of teacher-centered, lecture-based pedagogies in STEM (Fairweather, 2008; Freeman et al., 2014), traditional methods of teaching carry on as a main mode of instruction in postsecondary STEM education (Stains et al., 2018). Widespread uptake of student-centered instructional approaches has generally been slow (Laursen, 2019; Stains et al., 2018). That said, the proliferation of centers of teaching and learning aross universities (Kelley et al., 2017) has helped to engage more undergraduate STEM faculty in enacting alternative, student-centered practices to equitably teach subjects like mathematics (Reinholz, 2023).

This article accordingly reviews empirical research on instructional changes that faculty have attempted to make postsecondary STEM courses more equitable. Our research questions are as follows: How do interventions to advance equity in undergraduate STEM courses conceptualize and operationalize the study of equitable STEM teaching and learning? What prominent pedagogical frameworks do these interventions draw on, and how are the pedagogical frameworks applied to instruction as well as assessed for outcomes? By systematically reviewing empirical studies of undergraduate pedagogical interventions across disciplines within STEM published from 2010 to 2023, we aim to contribute an analysis of opportunities and gaps in equitable postsecondary STEM education scholarship and practice.

Review of STEM Education Research and Alternative Pedagogies

Since 2020, several systematic reviews on STEM education have been particularly helpful in shaping and summarizing trends in the field. First, in a review of 798 articles published between 2000 and 2018, Li et al. (2020) found that research in STEM education has not only increased significantly in quantity since 2010, it has also grown internationally in importance. Although the field has been dominated by authors from the USA who comprised 75% of the publications they reviewed, there is demonstrable interest in the topic especially from researchers in Australia, Canada, Taiwan, Spain, and the UK. Moreover, nearly half (48%, or 375) of the articles they reviewed were about “goals, policy, curriculum, evaluation, and assessment” related to STEM education, while about 13% (i.e. 108) of the studies were about “teaching, teacher, and teacher education” at the K-12 level. Only 18 articles in total fell under the topic category of teaching at the postsecondary level.

Li et al.’s (2022a) subsequent review of the one hundred most-cited empirical studies of STEM education published in 2000 to 2021 revealed that the field is trending towards an interdisciplinary perspective of STEM education, as opposed to earlier publications that were grounded within a specific discipline of STEM. Furthermore, they identified “culture, social, and gender issues in STEM education,” which accounted for nearly half (48) of the one hundred publications they reviewed, as the most popular topic within STEM education research. This was followed by the topic area of “STEM learner, learning, and learning environments at [the] postsecondary level.” In contrast to Li et al.’s (2020) general review which showed a dearth of STEM education research at the postsecondary level, this review of only high-impact articles found that 63% of publications focused on postsecondary-level issues. A later literature review conducted by Li et al. (2022b) of the most-cited articles from 2020 to 2022 confirms this trend of an increased focus on culture, social, and gender issues, along with postsecondary STEM learning.

In comparison with Li et al. and’s (2020, 2022a, 2022b) reviews of the general field of STEM education, Bottia et al.’s (2021) literature synthesis was more specific in examining only postsecondary STEM education and college students who are underrepresented in STEM. Their review of fifty articles sought to identify what factors can help to explain RM college students’ decreased participation in STEM majors and careers. They found that one such factor is less exposure to curricula and instruction that is inclusive of RM students. The use of innovative instructional techniques like active learning (Rainey et al., 2018), along with assignments that meaningfully connected to students’ lived experiences (Barker et al., 2014), were among the identified predictors of students’ success and persistence in college STEM. Yet they observed that RM students were often enrolled in “culturally unresponsive” classes that “do not accommodate the learning styles, cultural values, and preferences” of RM students (Bottia et al., 2021, p. 635).

Although Bottia et al.’s (2021) review is helpful in zeroing in on the general importance of curricula and instruction in advancing equity in STEM, its focus is not on presenting or summarizing the various ways by which faculty can be innovative in their instruction. Namely, besides active learning, there are a myriad of pedagogies that faculty can draw from to shape their course design. This ranges from collaborative learning, experiential learning, flipped learning, problem-based learning, universal design for learning, service learning, global learning, gamification, trauma-informed pedagogy, critical pedagogy, feminist pedagogy, translanguaging pedagogy, anti-racist pedagogy, to abolitionist teaching. This pedagogical decision is complicated by the fact that educational equity can take on a multitude of meanings, such as “…equal outcomes for every learner, [but also] equal educational experiences for each child, or equal levels of growth or development for each learner” (Levinson et al., 2022, p. 2). These varying conceptions of equity can in turn influence how faculty teach. For instance, Russo-Tait’s (2023) thematic analysis study found that professors who conceptualized equity as equality tended in fact to veer towards lecture-based instruction. This stands in contrast to professors who conceptualized equity as inclusion using active learning and/or inclusive teaching practices, and professors who conceptualized equity as justice adopting critical pedagogy as a guiding framework.

Of the multitude of pedagogical frameworks listed above, we briefly summarize three major frameworks that have guided many reforms in undergraduate teaching and learning: (1) active and experiential learning, (2) critical and culturally responsive pedagogy, and (3) universal design for learning.

Active and Experiential Learning

In contrast with a lecture-based, expository method of direct instruction, active learning is based on the theory of constructivism and the goal of promoting students’ participation in creating knowledge (Bonwell & Eison, 1991; Davidson & Katopodis, 2022; Freeman et al., 2014). Active learning benefits underrepresented students in STEM majors (Theobold et al., 2020), and comprises instructional activities that can range from using clickers or response systems to conduct polls, to having students role play situations or scenarios, to incorporating a group jigsaw activity. Because it rests on a general notion of engaging students in the learning process, active learning has become an umbrella term for many models of teaching and learning such as problem-based learning (Hsieh & Knight, 2008), project-based learning (Lehmann et al., 2008), inquiry-based learning (Laursen, 2019), context-based education (Gilbert, 2006), or team-based learning (Haidet et al., 2014; Najdanovic-Visak, 2017). Distinctions can be made between each of these models, such as how team-based learning—which makes small group work the primary in-class activity—differs from the simple incorporation of a think-pair-share or the requirement of accountability and roles in cooperative learning (Michaelsen et al., 2023). But all active learning strategies share the view of transforming teaching from the passive transmission and receipt of knowledge to an environment in which students are actively engaged in constructing their learning process.

Experiential learning, often called learning by doing or applied learning, is more specific than active learning in that it is structured around a process of engaging students in direct experience and focused reflection. Grounded in the theoretical work of David Kolb (1984), experiential learning entails guiding students through a feedback loop of four stages: concrete experience, reflective observation, abstract conceptualization, and active experimentation. Experiential learning does not simply mean to engage students in hands-on experiences, but also to reflect on that experience, and experiment by applying what they have learned from that reflection to a new situation.

Critical and Culturally Responsive Pedagogy

Similar to active learning, critical pedagogy shares the goal of decentering the role of the teacher to empower students in a process of agentic learning. However, critical pedagogy adopts a sociopolitical lens of education in maintaining “that the classroom, curricular, and school structures teachers enter are not neutral sites…[but are rather] politically contested spaces” (Kincheloe, 2008, p. 2). Grounded in the work of Paulo Freire (1970), critical pedagogy seeks to increase students’ critical consciousness through examining systems and structures of power, inequality, marginalization, and oppression. Just as active learning can be interpreted as an umbrella term for many other models of learning, critical pedagogy encompasses: anti-racist pedagogy, which focuses on challenging systemic racism and developing students’ awareness of their social position (Blakeney, 2005; Kishimoto, 2018); racially liberatory pedagogy which aims to emancipate students from from racist ideologies and oppression (Castillo-Montoya et al., 2019; reality pedagogy, which uses students’ interests and experiences to make science content in particular more meaningful and engaging (Emdin, 2011); and engaged pedagogy, in which teachers also reflect on their own growth and self-actualization in the process of empowering students (hooks, 1994).

Culturally relevant pedagogy is similarly founded on the importance of guiding students in their educational experiences to recognize and critique social inequities. Rooted in the pillars of developing students’ academic success, cultural competence, and critical consciousness (Ladson-Billings, 1995), culturally relevant pedagogy has evolved over the years to also be termed culturally responsive teaching (Gay, 2000) and culturally sustaining pedagogy (Paris, 2017). Gay emphasized the need to teach to and through students’ cultural knowledge and prior experiences, while Paris advanced the field to consider ethnic and linguistic differences not only as strengths but also as differences that should be actively maintained.

In line with Ladson-Billings and Gay’s identification of culturally responsive pedagogical practices in K-12 settings, scholars like Tuitt, 2003, 2016) have demonstrated how critical and inclusive pedagogy can be implemented in higher education through: (1) leveraging the lived experiences of historically marginalized students, (2) creating identity-safe classroom spaces, and (3) adopting interactive instructional practices that draw from interdisciplinary perspectives. While there is often buy-in that these are effective teaching practices, the need to maintain “academic rigor”can prevent instructors from implementing such pedagogy in higher education. To this end, Castillo-Montoya (2018) shows how academic rigor in higher education can be reconceptualized as “deep, inquiry- and equity-based learning” (p. 42) grounded in: connecting subject-matter content with the lived experiences of racially and ethnically diverse first-generation college students; and incorporating their funds of knowledge to further advance student learning.

Universal Design for Learning (UDL)

UDL amounts to a set of design principles and instructional approaches to help all students—not just those with disabilities—have equal opportunities to learn (Basham & Gardner, 2010). In creating a student-centered classroom and curricula, UDL principles underscore the need to first recognize learner differences so that an instructor may then adopt flexible approaches to teaching and learning (Rose & Meyers, 2006). These approaches are centered around a set of guidelines for accessible and inclusive instruction that “provide[s] students with choices and alternatives in the materials, content, tools, context, and supports they use” (Izzo & Bauer, 2015, p. 18).

Specifically, students are provided with multiple means of engagement, representation, and expression that respectively capture the “what,” “why,” and “how” of learning (CAST, 2018). For instance, providing students with multiple means of engagement can entail giving students choice and autonomy in assignments, increasing the use of mastery-oriented feedback to students, and helping students develop self-regulation skills through self-assessment and reflection. These practices help students become more self-motivated and purposeful in their learning process. Similarly, an instructor can provide students with multiple means of representation by offering students alternatives for visual or auditory information, supporting students in decoding text or mathematical symbols, and developing student comprehension through highlighting patterns and big ideas. These practices together help to develop resourceful and knowledgeable learners. Finally, an instructor can cultivate strategic, goal-directed learners by providing multiple means of action and expression, which can range from increasing students’ access to tools and assistive technologies, using multiple media for communication, and guiding students in setting appropriate learning goals. In general, there is overlap with active learning in that UDL principles also encourage the use of active learning methods when offering students multiple means of engagement and expression (Brusini et al., 2019; Sasson et al., 2022).

Methods

This systematic review adhered to guidance provided by Siddaway et al. (2019).

Search Terms and Databases

The search strategy was designed to capture studies that examined applying at least one of nine target pedagogical frameworks (i.e., culturally responsive pedagogy, reality pedagogy, critical pedagogy, translanguaging, active learning, context-based education, UDL, experiential learning, and problem-based learning) to address racial or gender inequities in undergraduate STEM education. Given the vast amount of pedagogical frameworks that exist, often referring to similar conceptual frameworks, we constrained our search to only these target frameworks that we identified at the outset of the study. We then developed search strings—each maintaining identical keywords and subject headings for undergraduate STEM education, diversity, and inclusion, while varying the keywords and subject headings for each of the nine target frameworks (see Appendix 1).

The search was executed on September 8, 2023, in the Education Resources Information Center (ERIC) via EBSCO database. Similar to Li et al. (2020) and Li et al., (2022a, 2022b), we initially restricted our search to articles published after the year 2000. The nine searches collectively yielded 809 citations (see Table 1). Searches for reality pedagogy and context-based education both returned 0 records and were eliminated from the study. After loading the citations into RefWorks citation manager, 56 duplicates were identified, resulting in 753 unique items. The outcomes of our search, de-duplication, and screening processes are presented in the PRISMA flow diagram (see Fig. 1).

Table 1 Total records per target framework from initial database searches
Full size table
Fig. 1
figure 1

PRISMA diagram of searching and screening. This figure is adapted from Page et al. (2021)

Full size image

Screening Process

Initial Coding

We first reviewed the title and abstract of the remaining 753 articles to inductively code each record by the: disciplinary field (i.e., agricultural life science, astronomy, biology, chemistry, engineering, geosciences, mathematics, medicine, physics, physiology, social science, or general STEM course); methodology (i.e., empirical or non-empirical); equity focus (i.e., race, gender, disability, language, rural, income, first-generation, community college); and outcome (i.e., achievement/content knowledge, anxiety, STEM appreciation/interest, belonging/support, engagement/motivation, retention, self-efficacy, and social justice/critical consciousness). Our inductive codes for equity focus areas notably expanded during the search process due to the emergence of studies that went beyond race and gender disparities to consider inequalities in STEM education faced by other student subgroups, such as students with disabilities, first-generation college students, and community college students—the latter of whom are more likely to face additional challenges in higher education like the need to work, being part-time students, and requiring developmental education (Varty, 2022).

Inclusion and Exclusion Criteria

We then reviewed article abstracts to apply an inclusion criteria consisting of the following: (1) articles from peer-reviewed journals, (2) articles written in English, (3) articles with a clear reference to one of the identified pedagogical frameworks, (4) studies of an undergraduate STEM course (in accordance with the NSF STEM education course designations which includes courses in social, behavioral, and economic sciences), and (5) empirical studies that included quantitative, qualitative, and mixed methods of research. Although we did not actively search for studies situated in the field of scholarship of teaching and learning, such studies were included so long as it included empirical research methods that extended beyond reflections of teaching (Kern et al., 2015).

We excluded citations that were books, book chapters, reports, and conference proceedings because we could not be assured that such citations underwent a peer review process. Since the concept of equity is essential to our systematic review, we further refined and revised our initial cutoff start date from 2000 to 2010 based on Goad and Chartwell’s (2022) analysis showing how diversity, equity, and inclusion (DEI) related language across STEM fields in US universities has increased by over 4000% since 2010. We also excluded publications that were not related to STEM pedagogy, such as articles on STEM cell research, studies of non-STEM courses like education or arts courses, or studies of STEM programs that were not pedagogical in nature (e.g., mentoring, undergraduate research, or internships). Publications that did not focus on teaching and learning with students, such as articles solely about faculty professional development, research on teaching assistants, and autoethnographies of STEM instructors, were likewise excluded. Articles that focused on doctoral or graduate students, elementary or secondary education, or a combination of secondary and undergraduate education were excluded as well. Additional exclusion criteria included: records that made no reference to increasing equity, inclusion, or diversification in the abstract; and articles that were not an empirical study of an intervention (e.g., illustrative case study of applying a framework, review of other studies, theoretical paper, or essay).

Data Analysis

This first screening phase of applying our inclusion and exclusion criteria to reviewing the title and abstracts of our records narrowed down our search to 41 articles (see Fig. 1). We then downloaded and reviewed the full text of these 41 articles to conduct an initial round of analysis to surface common themes (Saldaña, 2015). We took analytic notes summarizing how equity was conceptualized in the study, how the pedagogical framework was applied in a classroom intervention, and what the general findings were. We also created additional codes for the discipline and geographical location of the study, as well as sub-codes for the pedagogical framework (e.g., inquiry-based learning that falls under active learning) and research methods (e.g., ANOVA of gender differences on survey vs. semi-structured interviews).

The second screening phase of reviewing the full text led us to further remove articles that were: research briefs; not a study of enacting equity (e.g., though it took place in a minority serving institution); not a study of an alternative pedagogical framework (e.g., study of online learning, or an intervention of learning student names); and not an empirical study of an intervention (e.g., illustrative case studies, design studies, teaching reflections, or pilot studies). We did, however, retain empirical research that were not explicit studies of the effects of an intervention but were interview studies about student perceptions of a particular pedagogical framework or intervention. This ultimately yielded 31 studies for our final review (see Appendix 2).

We then reviewed these 31 studies and again took another round of more specific analytic notes on how equity was framed and operationalized, how the alternative pedagogy was applied, and what the outcome construct and direction of the findings were. This helped us to inductively identify final themes and codes aligned to our research questions, which are explained in our reporting of the results below.

Results

Before presenting our results below organized by research question, we first provide a basic summary of the geography and discipline of the 31 research articles. All but two of the 31 studies took place in the USA; the two exceptions were in the UK and Switzerland (see Fig. 2). Nearly a third (nine out of 31) of the studies were conducted in the Southeast of the USA, followed by the West and Northeast (four studies each). Five studies notably were held at multiple universities across the USA, suggesting the occurrence of some collaborative efforts to implement and research STEM instructional reforms at a broader scale.

Fig. 2
figure 2

Frequency of studies by geography

Full size image

Our results also show that most (nine out of 31) pedagogical reforms and studies were conducted in undergraduate biology courses (see Fig. 3). Almost 20% (six out of 31) of the studies were of mathematics courses, followed by physics and general STEM courses (four out of 31 each). Interestingly, only two studies were of chemistry or engineering classes, and only one study took place in a computer science course.

Fig. 3
figure 3

Frequency of studies by discipline

Full size image

  • RQ1: How Do Interventions to Advance Equity in Undergraduate STEM Courses Conceptualize and Operationalize the Study of Equitable STEM Teaching and Learning?

Our findings show that most studies (55% of the 31 studies) conceptualized equity as an underrepresentation of a particular group based on race, gender, income, disability, and/or first generation in college status (see Table 2). Almost 30% of studies framed equity as closing or reducing an achievement or performance gap, although four out of the 31 studies focused on mitigating disparities in self-efficacy or confidence. Four studies referenced systemic inequities, such as women having less access to resources and support, or the unique stressors that community college students often encounter in persisting and successfully transferring to four-year programs. Two studies specifically framed these challenges as sociocultural barriers that certain groups face, like RM students having to overcome deficit perspectives and low expectations. Finally, two other articles described equity as issues of inclusivity, while one study framed equity in general terms of educating diverse student populations.

Table 2 Conceptions and operationalizations of equity
Full size table

In terms of methodologically operationalizing an analysis of equity, we found that the vast majority (22 or 71%) of the 31 studies tended to compare demographic groups (see second section of Table 2), such as looking for differences in performance or self-efficacy between men and women or RM and non-RM students. Four out of the 31 studies examined equity by narrowing the study design and participants specifically to a marginalized group such as community college, Historically Black Colleges and University (HBCU), or Lesbian, Gay, Bisexual, Transgender, Queer, Intersex, and Asexual (LGBTQIA) students. Sixteen percent of the studies, however, only incorporated equity in framing the study, but not in the analytical plan and results.

We then further analyzed the 26 studies that operationalized equity to understand their focus in analyzing disparities or inequities among student groups. We found that twelve (or 46%) of the studies focused on gender differences, while one study evaluated differences between gender as well as student classification by year of study. Five of the 26 studies analyzed racial differences, while two studies examined race and gender differences. Two studies looked at outcomes for first generation college students—one in combination with students with low income, and the other in combination with race and gender. Two studies selectively focused on improving outcomes for community college STEM students, while one study each focused on improving outcomes for LGBTQIA STEM students and STEM students attending HBCUs.

  • RQ2: What Prominent Pedagogical Frameworks Do These Interventions Draw on, and How Are the Pedagogical Frameworks Applied to Instruction as well as Assessed for Outcomes?

Of the 31 studies, all but two drew on active learning as a general framework for the classroom intervention. Some studies specifically referenced theories like constructivism or Bandura’s theory of social learning, while others connected the study to project-based, inquiry-based, or team-based approaches to learning. Despite the varied approaches, these studies all shared the common theme of disrupting a traditional lecture-based approach to teaching in order to engage students in their learning process. Instead of active learning, two studies grounded themselves in culturally responsive pedagogy.

Considering the wide spectrum in how active learning or culturally responsive pedagogy can be implemented, a deeper analysis of how these studies applied these frameworks to instruction showed that over 80% of studies incorporated group or peer collaboration into course meetings or assignments (see Table 3). Over 40% specifically integrated more discussions into the class, while about one-quarter used student-response systems like clickers to poll students during lessons. Five out of the 31 studies referenced adopting a flipped classroom approach to teaching, where students learn content knowledge before class and focus on applying that knowledge in class. The same number of studies described incorporating hands-on activities and systems of feedback into the class. Three studies explicitly mentioned changes to course assessments and grading as part of the intervention, while two and one study respectively mentioned integrating growth mindset and metacognitive reflection into the course. Finally, the two studies that used culturally responsive pedagogy instead of active learning emphasized having students learn about and/or engage in assignments connected to their backgrounds aimed at increasing their critical consciousness.

Table 3 Summary of pedagogical strategies applied
Full size table

Lastly, we examined the 31 studies to look for trends in what outcomes they analyzed, as well as the directionality of the outcomes (see Table 4). At almost one third of the 31 studies, the most prevalent outcome we identified was reducing a performance gap in test scores or grades. Almost one-quarter (i.e., seven or 23%) of the studies looked for effects of the intervention on students’ self-efficacy. In contrast with reducing a gap in grades between different demographic groups, four studies assessed the outcome of increasing grades or content mastery for all students. Interestingly, four studies also focused solely on differences in participation or communication as an outcome. Other outcomes included anxiety, belonging, motivation, perceptions of active learning, and reducing a gap in perceived learning (as opposed to actual scores or course grades).

Table 4 Cross-tabulation of study outcomes and directionality
Full size table

Our cross-tabulation of outcomes and directionality showed that 24 studies found positive effects from the intervention, compared to five, one, and two studies that found negative, no, and mixed effects respectively (see Table 4). Although empirical in nature in terms of using quantitative survey or qualitative interview methods, seven of the studies did not seek to assess the impact of the intervention. These studies instead focused on describing student experiences or factors that influence anxiety or belonging in the context of active learning courses.

Discussion, Implications, and Limitations

This study aimed to examine empirical trends in how undergraduate STEM faculty have attempted to advance equity through instructional changes. We first found that most studies are concentrated in the USA within the field of biology, followed by mathematics and then physics. These results show opportunities for more research on implementing equitable pedagogy in undergraduate STEM courses particularly in the disciplines of chemistry, engineering, and computer science. Future research can also engage in the broader scholarship of teaching and learning (Kern et al., 2015) across STEM disciplines, rather than taking disciplinary based approaches to education research. Our findings additionally suggest the field can be strengthened by more studies conducted at multiple university sites to understand the effects of implementing alternative pedagogies of equity at scale.

We also found an emphasis on examining gender equity first and foremost, followed by race. It was noteworthy that we had to expand our search to include other identities besides race and gender, but there were substantially fewer empirical studies that focused on other demographics such as students with low income, first-generation college students, students who are not heteronormative, and students with disabilities. Future research can pay specific attention to these understudied student subgroups. Furthermore, students do not typically experience these identities in isolation. Adopting an intersectional focus—of examining race, gender, and disability status as an example—not only honors students’ multiple identities, it affords a deeper understanding of how these factors interact to affect students’ STEM learning as a member of multiple minoritized student subgroups.

Our findings also show a tendency towards conceptualizing equity as an underrepresentation of a particular subgroup of students, along with operationalizing an empirical study of equity as comparing and reducing a performance gap between demographic groups. This corroborates Ives et al.’s (2023) finding that most federally-funded initiatives to diversify undergraduate STEM used the outcomes of retention, graduation, and cognitive achievement measures to assess program impact. Our results could be a product of this systematic review’s focus on only including empirical research on equity in STEM instruction. Although our search initially included more articles that, for instance, framed equity in terms of inclusivity, many of those articles were omitted during the exclusion process because of its non-empirical nature.

But considering Levinson et al. (2022) and NASEM (2022) drawing attention to the multiple ways by which educational equity can be defined, this finding also brings into relief the potential consequences of not conceiving of equity in other ways. Many scholars have argued that a narrow focus on the achievement gap produces myopic, short-term solutions at best (Ladson-Billings, 2006). Worse yet, it can further perpetuate dehumanizing stereotypes of Black and Latinx students’ academic performance (Gutiérrez, 2008). Moreover, Ives and Castillo-Montoya’s (2020) systematic review of literature on first-generation college students found a tendency to position first-generation students against normative ways of learning that ultimately perpetuates an assimilationist view of what it means to be an academic achiever and learner. In contrast, they found fewer studies that conceptualized first-generation college students as learners whose lived experiences “can contribute to their academic learning, advancement of disciplines, self-growth, and community development” (p. 168). Similarly, we argue that an overemphasis on representation within the STEM equity discourse can inadvertently veer the field towards checking diversity boxes, meeting RM or gender quotas, or maintaining the status quo of incorporating and assimilating minoritized students into a culture dominated by males and Eurocentric conceptions of STEM. Even when indigenous STEM knowledge is incorporated into redefining STEM content or adopting a multicultural approach towards teaching science and mathematics, there is a risk of reductively romanticizing minority cultures and increasing minoritized students’ “otherness” (Eglash, 1997). A one-dimensional view of equity can similarly reduce well-intended STEM equity initiatives to assimilationism, tokenism, and cultural essentialism.

Within the broader societal discourse on DEI, Davis (2021) has argued that DEI initiatives have persistently failed to move the needle because diversity, equity, and inclusion conceptually rests on notions of representing and including minoritized groups into existing dominant cultures that are riddled with inequalities. Instead, she calls for actualizing the values of belonging, dignity, justice, and joy (BDJJ) because such a shift would lead to a different outcome of redefining communities with the voices of marginalized people at the center. Accordingly, future STEM research can expand beyond assessing the representation of minoritized students or reducing a performance gap between demographic groups to study constructs of joy or belonging as outcomes (see Hansen et al., 2023; Pai, 2023; Smith & Lucena, 2016; Yeager et al., 2016), including empirical examinations of students’ qualitative experiences of equitable STEM education (see Flores et al., 2023). In line with Davis’ (2021) BDJJ framework, more STEM education research and practice can also focus on increasing students’: sense of belonging, experiences of feeling valued, and opportunities to examine power and systemic oppression in STEM. This would additionally help the field in approaching NASEM’s (2022) latter two equity goals of expanding what constitutes STEM and seeing STEM as part of broader justice movements.

In terms of pedagogical frameworks, our analysis shows an overwhelming trend towards using active learning. Our objective of being more nuanced in teasing out how interventions applied an alternative pedagogy revealed that incorporating groupwork and peer collaboration were at the center of many instructional interventions. This begs the question of what precisely is it about discussions and collaboration that fosters equity for students, and how should discussions and collaborative work be formatted? While it is not possible to definitively answer this question through this systematic review, two of the studies we reviewed suggest it might specifically be the opportunity for students to learn from multiple perspectives that creates more equitable learning experiences (Chen et al., 2015; Lewis & Estis, 2020).

While the use of active learning as a framework represents a progressive shift in STEM education, there are potential drawbacks to applying frameworks not explicitly designed for RM students to transform the outcomes of RM students. Ives and Castillo-Montoya (2022) argued that “when higher education scholars use frameworks and theories that encourage assimilation of marginalized students, their findings will follow suit” (p. 163). For instance, they discuss how applying frameworks like Bourdieu’s notion of social and cultural capital are no doubt contributions to the field, but defining cultural capital based on the dominant culture can also represent a failure to recognize nondominant forms of social capital. Similarly, we argue that while active learning pedagogy can engage students in deeper STEM learning, it is not explicitly designed for RM students. Namely, it does not make use of or elevate racially minoritized students’ funds of STEM knowledge. Nor does it help students to see the relationship between STEM education and broader social justice.

To this end, critical and culturally responsive pedagogy is specifically designed to transform educational experiences for RM students, yet we found that few studies adopted a critical, socio-political lens to reforming pedagogy. We theorize this could be for two reasons. First, whereas adopting active learning methods entails a change to how the class is taught, adopting critical or culturally responsive pedagogy might necessitate a change in content to what is being taught. For example, course assignments might have to be changed towards examining issues of social inequities, which may be a barrier for faculty who may already feel the pressure of covering a large amount of course content. Second, Shultz et al. (2022) found that many STEM faculty adopt a culture-free lens of their discipline, such as the notion that mathematics is universal. This disciplinary perspective makes it less likely for faculty to enact culturally relevant pedagogy in their courses. That said, they did find cases of instructors being able to hold contradictory epistemological beliefs, namely in being open to practicing culturally relevant pedagogy despite believing their discipline is culture-free. We thus argue that there are opportunities for future practice and research on equitable STEM pedagogy to draw on critical pedagogy in order to cultivate and sustain students’ cultural competencies and sociopolitical consciousness.

Lastly, it is difficult to ignore the overwhelmingly positive directionality of the research findings we reviewed. While we do not question or negate these positive results, this observation may indicate room for more critical or nuanced analyses of the challenges associated with enacting equitable STEM pedagogy. Notably, five studies did critically surface negative findings, with three out of these five studies all examining the outcome of equity in participation between men and women. This further suggests the possibility that while there may be progress towards equity as defined by the outcome of reducing gaps in test scores and course grades between demographic groups, there may still be inequities in mediating factors like participation and self-efficacy that mark students’ learning processes.

Limitations

Since this review is limited to only empirical studies of classroom interventions, it is important to note that our findings should not be conflated with what is widely being applied in postsecondary classes. For instance, while there was a higher frequency of studies in our review within the discipline of biology, this should be not be misinterpreted as biology courses and instructors being the ones within STEM disciplines to make the most attempts at transforming teaching and learning away from traditional lecture-based instruction. Since this study excluded non-empirical studies, it is possible that instructors from other disciplines apply alternative equitable pedagogies, but are simply not conducting empirical research and publishing on those pedagogical changes in journals that met our inclusion and exclusion criteria. Similarly, active learning was most popularly applied according to our review of the empirical research, but this should not be conflated with the fact that undergraduate faculty are not at all applying other pedagogical frameworks like critical pedagogies and UDL. In fact, it is important again to point out that there were initially more articles, for example that applied UDL frameworks to accommodate students with disabilities in STEM, that were excluded from the final review because they were not empirical in nature.

We also recognize that the empirical studies we reviewed were constrained to the ERIC database, which skews heavily towards US research. This could have contributed to why only two of the articles we reviewed were from outside of the USA. Future reviews can consider casting a wider net to focus on and/or compare international research on equitable pedagogies applied in higher education settings across the world.

Conclusion

This systematic review aimed to contribute an overview of how higher education faculty are enacting and researching instructional interventions to make postsecondary STEM courses more equitable. We found that most of the empirical research we reviewed were based in the USA and fell within the disciplines of biology and mathematics. Studies also largely applied active learning centered around promoting group and peer collaboration, rather than applying other pedagogical frameworks such as culturally responsive pedagogy or UDL aimed at increasing students’ critical consciousness or providing multiple means of expression, representation, and action. We also found a propensity towards examining gender and racial equity, and towards assessing the outcome of reducing the performance gap between over- and under-represented demographic groups with largely positive results. These findings suggest opportunities for more research focusing on other subgroups such as students with disabilities or low income, as well as research that departs from the conventional DEI framework to analyze other outcomes like belonging, dignity, justice, and joy (Davis, 2021; see also NASEM, 2022).

Our analysis of how studies conceptualize and operationalize equity also suggests the STEM education discourse is dominated by framing equity as the underrepresentation of particular subgroups of students and comparing gaps primarily in performance. This convergence mutes a discussion of larger questions such as the following: Who defines equity? Is there cognitive justice in how the concept of equity is defined within the STEM education field? Are definitions of equity being pushed from the top down, or bottom up with the inclusion of faculty and student voice? Moreover, while transforming pedagogy is core to improving STEM education, it does not in and of itself assure equity in and across the STEM field. In other words, pedagogy alone cannot induce system-wide equity. The larger goal of attaining equitable educational outcomes requires the creation of conditions that allow for equity to emerge. It also requires the provision of sufficient resources and support at various levels beyond just a class; these levels range from a department, to school, to college, to university system. Future research can correspondingly adopt a more expansive view of the ways in which pedagogy interacts with the other institutional conditions that are necessary to foster STEM students’ learning.