A Commentary on the Upending of DEI Research in STEM: A Testimonio on Grant Termination in the Sociopolitical Context

Abstract

Recent actions in the U.S. have sought to curtail diversity, equity, and inclusion (DEI) efforts through policies that restrict DEI programs and offices and through the cancelation of research grants. In this commentary, I offer a testimonio to document how my federally funded project on racial microaggressions in STEM was defunded. The testimonio approach allows for a first-person account of lived experiences to address inequity embedded in social and political realities. I describe the nature of my project, the antecedents and consequences of the grant cancelation, and broader implications. Additionally, I address the sociopolitical context of grant terminations by the current administration and potential responses by our STEM community.

1. Introduction

This Special Issue on Creating Cultures and Structures of Opportunity in STEMM Ecosystems comes at a precarious time for research on equity in STEM education. Recent actions in the US have sought to curtail diversity, equity, and inclusion (DEI) efforts through funding cuts and policies that restrict DEI programs and offices (The Chronicle of Higher Education, 2026). The primary aim of this commentary is to document the conditions under which federally funded research on STEM inclusion was abruptly canceled by the federal government last year in compliance with Executive Order 14173 (National Science Foundation, 2025). Specifically, I will describe how my project on racial microaggressions in STEM settings was funded and defunded by the National Science Foundation (NSF). The objective of documenting this instance of STEM research termination will be met by addressing: (1) the sociopolitical context in which the grant termination occurred, (2) the nature of the research that was terminated, and (3) potential responses by the STEM community. Documenting this termination is significant for several reasons: (a) to record its occurrence in the timeline of STEM education for historical reference; (b) to attest to preceding events and consequences; and (c) to strategize for persevering in STEM inclusion research.

Additionally, I argue that these current events demonstrate that, more than before, vigilance is needed to monitor whether policy reflects scientific knowledge or acts against it. Although this vigilance is typically associated with science on climate change, technology, or medicine, it applies to social and education sciences as well. As signaled by this Special Issue, much is known about how to create and expand opportunities in STEM because of social and education sciences. We, as scholars, are positioned not only to make policy recommendations but also to call out policies that run contrary to research.

For this commentary, I adopt a testimonio (testimony) approach to document the termination of my federal research grant on racial microaggressions in STEM. Testimonio refers to a first-person account, rooted in Latin American cultural traditions and socio-political histories (Cervantes, 2020; Delgado Bernal et al., 2012). Testimonio is a multifaceted approach that differs from typical narratives or autobiographies in its intention to raise awareness of injustices and promote solidarity with a marginalized community (Cervantes, 2020; Comas-Díaz, 2020; Delgado Bernal et al., 2012; García-Díaz, 2025). The use of testimonio among academics is diverse and growing (García-Díaz, 2025), with a common focus on gathering narratives and experiences as sources of lived knowledge and responses to inequity (Cervantes, 2020; Delgado Bernal et al., 2012). Recent examples of testimonio in STEM education research include studies on the experiences of Hispanic/Latine1 preservice teachers (Chapman et al., 2026) and engineering students (Mejia et al., 2022). Testimonios are cultural perspectives marked by an “urgency and intentionality” in “advocating for educational rights” (García-Díaz, 2025, pp. 744, 746).

In proceeding with this testimonio on the government-directed cancelation of a federally funded STEM education research project, I identify the boundaries of this commentary. This is not a research report that details traditional research questions and methods of data collection and analysis. However, I describe the project I proposed on racial microaggressions in STEM to document the nature of research targeted for termination. This article is not a policy analysis because its primary aim is to document an unprecedented event in funding for higher education research (Mervis, 2025a) from the personal perspective of the author. I will address the sociopolitical context of the event and the link between science and policy in expanding STEM opportunities, drawing upon my experiences. Finally, this article is not a review of STEM-DEI or science-policy literature. I cite relevant literature to fortify arguments and perspectives, using the testimonio approach of speaking from lived experiences through “sociopolitical realities” (Delgado Bernal et al., 2012, p. 364). Within this scope, I will address the three points identified in the article’s opening.

2. Sociopolitical Contexts of Opportunity, Research, and Termination

To begin this testimonio, I describe the sociopolitical contexts in which my STEM DEI research evolved, leading up to the grant termination. I start by discussing my academic background to situate events within their historical and sociopolitical context. In the 1970s, my father used the G.I. Bill to attain an associate’s degree at the local community college. At that campus, a few miles from the U.S.-Mexico border, a Chicano Rights-atmosphere buoyed my father’s notion of boundless educational opportunities for his children. In the 1990s, I was admitted to a prestigious university due to my parents’ belief (impacted by the U.S. Civil Rights Movement) that affirmative action policies and Pell Grants could open doors for me as a high school valedictorian whose family could not afford a four-year degree. Social and educational policies shaped my family’s beliefs about access to education for underserved groups.

Against the backdrop of challenges to affirmative action (e.g., California Proposition 209, 1996; Gratz v. Bollinger, 2003), I became the first in my family to earn a bachelor’s degree and advanced to graduate school. In transformative experiences, I was prepped for graduate school by a Summer Research Opportunities Program (SROP) and funded by a Ford Foundation Fellowship, which (at the time) were aimed at students of color. Incidentally, the Ford Fellowship Program was discontinued in 2024 (see announcement by former president Walker, 2022), and SROP applicant criteria have since been broadened beyond racial/ethnic minority status (see, for example, Rackham Graduate School, 2026).

My doctoral training in social psychology occurred in the late 1990s to early 2000s, during a period of intense study of implicit cognition, before the terms “implicit prejudice” and “unconscious bias” entered mainstream vernacular. At the time, social psychologists noted that blatant prejudice and explicit stereotyping had become less socially acceptable in modern American society, so research turned toward studying more subtle or implicit forms of bias (Sekaquaptewa et al., 2003). With my advisor, we focused our research on stereotypes and the extent to which they are automatically activated and influence behavior (Sekaquaptewa et al., 2003; Sekaquaptewa & Espinoza, 2004).

As my career turned toward education research, first as a program director and evaluator, and then as a faculty member, opportunities to research STEM education abounded. National policies and initiatives (e.g., No Child Left Behind Act, 2002; Every Student Succeeds Act, 2015) directed research and funding toward improving STEM education for reasons ranging from economic prosperity and national security to cultural enrichment (National Academies of Sciences, Engineering, and Medicine, 2025). I applied a social cognitive approach to STEM education by studying the experiences of underrepresented groups in STEM. My scholarship was influenced not only by the surge of research on implicit bias and STEM education but also by the sociopolitical events that impacted my pursuit of higher education.

The peak of the COVID-19 pandemic highlighted women’s labor in the workplace and the household, and concurrent protests against systemic racism heightened DEI consciousness (Rohlinger & Meyer, 2024; García-Louis & Reyes-Barriéntez, 2022). During this period, I engaged in research and service dealing with supporting women faculty and staff (P. Espinoza et al., 2026), and I began work on racial microaggressions in STEM settings. Between 2022 and 2024, NSF solicited concept papers and grant applications for a new funding opportunity, Racial Equity in STEM Education, which was well-suited to my research.

By the time I was drafting my grant proposal, public concerns over COVID-19 and racism had abated, and movements to suppress DEI in academia ascended. Texas introduced and passed S.B. 17 to eliminate DEI offices and training (Texas S.B. 17, 2023). When serving as a faculty senate representative, I expressed concerns to university- and system-level administrators about the legislation: how can we discuss the implications of DEI research if we cannot implement DEI programs? Most responses consisted of gratitude for the bill’s “carve-outs” for research (As an aside, separate legislation a couple of years later, Texas S.B. 37 (2025), dissolved the faculty senate in public universities.).

In August 2004, I received funding for my NSF proposal to study microaggressions experienced by Black and Hispanic/Latine college students in STEM settings. In January, following the transition in federal executive power, I was notified that the annual meeting for grant recipients was canceled. Then, news broke of the mass cancelations of National Institutes of Health (NIH) research grants (Oza, 2025). My research team wondered if our NSF grant had escaped a similar fate, with our annual meeting as the only casualty.

For months, my collaborators and I had been meeting with student research assistants, dissecting research on racial microaggressions, and preparing materials for data collection. In exploring this topic, we shared our individual experiences with racial microaggressions. As the students described interpersonal exchanges—some of which they said had never been shared before—a few surprised themselves when they teared up. I believed that students’ vulnerability in describing those events would be recompensed through our research demonstrating how to counter microaggressions in STEM environments.

Then, in the spring, I received a notice from NSF stating that my study was not “in alignment with current NSF priorities” and was being terminated. The line in the notice that struck me most was: “NSF is issuing this termination to protect the interests of the government.” From one day to the next, our research assistants could no longer log hours, and the resources for data collection were revoked “to protect the interests of the government.”

My institution appealed the decision, and NSF responded that the appeal did not apply to its shift in priorities. Internal work was initiated to fund research assistants through the end of the term and help them find comparable positions. University administrators and department chairs reassessed budgets and time allocations. I informed project evaluators and other collaborators that meetings were canceled. I was required to submit a final project report in which customary questions regarding “impacts” and “accomplishments” were ill-suited for the circumstances.

Now, a year later, a variety of legal actions have led to the reinstatement of many NIH grants (Kozlov et al., 2026) and the reinstatement of some NSF grants identified in a lawsuit by University of California researchers (Thakur v. Trump, 2026; Mervis, 2025c), while other litigation continues, including a suit brought by the American Association of Colleges and Universities, the American Association of University Professors, and others (American Association of Physics Teachers, Inc. v. National Science Foundation, 2025). Despite court orders, it is unclear how much funding has been released to scientists (Kozlov et al., 2026). On the NSF website, a spreadsheet lists the terminated grants (see National Science Foundation, 2025, FAQs). In the list of 1667 entries, mine is listed as number 1459. I cannot help but liken it to a booking number, as if the project is guilty of endangering “the interests of the government,” to paraphrase NSF’s email notification.

An analysis showed that NSF grant terminations “reduced the diversity of NSF’s pool of funded scientists” (Mervis, 2025b, p. 809). Among terminated grants, 58% were led by women and only 35% by men2. However, men lead 60% of all active grants. Similarly, White scientists hold 60% of active grants but led about half of the canceled projects, while Black scientists lead only 4% of projects and held 17% of canceled grants. The analysis also showed primary investigators with disabilities and those who are Hispanic/Latine lost grants at twice their representation in the pool of active grants. A likely explanation for these findings is the tendency for researchers from underrepresented groups to conduct research involving STEM inclusion (Mervis, 2025b). This is certainly true in my case.

My guiding interest has been in how exclusion and equity are influenced by people’s beliefs and attitudes, as well as by social environments. In the next section, I describe my terminated project to develop a tool to counter racial microaggressions experienced by Black and Hispanic/Latine students, applying my social-psychological training to enhance the inclusivity of college STEM settings.

3. Countering Racial Microaggressions as STEM DEI Research

I provide a general discussion of my terminated research project, not necessarily to delve into the study’s technical aspects, but to share the nature of the research targeted by the current administration. As I will explain, the multi-year project was proposed to develop and test an innovative intervention for racial microaggressions in STEM, and it was discontinued in Year 1, prior to data collection. I describe this research to consider the implications of defunding and devaluing DEI research.

Among the barriers to STEM participation and retention for minoritized students is interpersonal racial bias in STEM settings, where Black and Hispanic/Latine students are subject to stereotypes of underachievement (Eaton et al., 2020; Flores et al., 2024; Leath & Chavous, 2018; Lee et al., 2020). A form of racial bias occurring in interpersonal interactions is racial microaggressions, which are stereotype-based verbal or nonverbal acts (Osman et al., 2024). These acts communicate exclusion or inferiority (Osman et al., 2024) rooted in stereotypical and racist perceptions of social group members. Racial microaggressions have been widely documented in the academic literature (e.g., Lui & Quezada, 2019; Nadal, 2011; Solorzano et al., 2000; Sue et al., 2007; Wong et al., 2014), and research on this topic continues to expand, including in relation to STEM settings (Ayala & Young, 2022; Hall et al., 2023; Moors et al., 2022; Osman et al., 2024; Smith et al., 2023).

Research on racial microaggressions in college settings points to their prevalence (see Ogunyemi et al., 2020, for a review). Students experiencing racial stressors in academic environments, including microaggressions, exhibit lower academic engagement and diminished institutional belonging (Durkee et al., 2021; Leath & Chavous, 2018). Research focusing on STEM in higher education has yielded similar findings. Underrepresented minority students who strongly identified with their domain of study and experienced frequent negative racial experiences were less likely to persist in their science major (Chang et al., 2011). STEM students who described experiencing racial microaggressions from faculty advisors expressed diminished academic efficacy, threatening their persistence in STEM (Burt et al., 2019). Experiencing racial discrimination by a professor was negatively associated with STEM persistence for Hispanic and Black undergraduates responding to the National Longitudinal Survey of Freshmen (Park et al., 2020).

More broadly, the consequences of experiencing interpersonal racial bias and microaggressions include decreased well-being, mental health, and physical health (see Choi et al., 2022, for a review), as well as depletion of cognitive resources needed for academic performance (Banks & Landau, 2021). Moreover, some scholars hold that racial microaggressions themselves are a form of oppression that sustains systemic racism (Skinner-Dorkenoo et al., 2021; Williams et al., 2020), as they reinforce stereotypes and perpetuate exclusion and power differentials between groups.

Given the consequences of racial microaggressions, I focused on developing a method to reduce them in college STEM settings. Current strategies for reducing microaggressions include preparing individuals to respond when microaggressions are directed at them (Sue et al., 2019) and implementing workshops that require substantial time and resources (Moss-Racusin et al., 2018; Williams & Gran-Ruaz, 2023). Further, the conundrum exists where people tend to respond more favorably to voluntary than to mandatory anti-bias training, yet individuals who could benefit most from the training avoid attending (Carter et al., 2020). Reviews and meta-analyses of racial bias interventions have yielded mixed findings regarding their efficacy (Bezrukova et al., 2016; Paluck et al., 2021; Röhner & Lai, 2021), although few interventions specifically focused on microaggressions have been empirically evaluated and linked to behavior change (Osman et al., 2024).

The approach I proposed for addressing racial microaggressions in college STEM settings situates the locus of change with those committing them, namely, STEM educators. Also, the format of this racial bias intervention differs from others and is more akin to communication used to promote public health or environmentalism. Public messaging, which seeks to influence smoking cessation, contraceptive use, or energy conservation, for example, draws upon psychological theories for influencing behavior change (Glanz & Bishop, 2010; Kusmanoff et al., 2020; Pope et al., 2018; Prochaska et al., 2008), such as the Theory of Planned Behavior (Madden et al., 1992) and the Information–Motivation–Behavioral Skills Model (Fisher & Fisher, 1992), among others.

The proposed tool, STEM Educator Series on Student Interactions (SESSI), utilizes these psychological theories of behavior change to convey visual and print messages intended to reduce racial bias in college STEM settings (P. Espinoza & Canaba, 2024). The series depicts students’ experiences of racial microaggressions in STEM settings (e.g., labs, classrooms, meetings) committed by STEM educators (e.g., faculty members, instructors, advisors). Importantly, each scenario in which a student experiences a microaggression from a STEM educator is illustrated beside an alternative scenario in which the microaggression is substituted with inclusive behavior. In other words, not only is an example provided of how a student experiences a racial microaggression in a specific STEM setting, but also of how an educator can instead enact an inclusive behavior in that setting. Inclusive behaviors such as these are referred to in the literature as microaffirmations: acts that convey belonging and validate others who may be excluded or ignored based on their racial/ethnic identity (Ellis et al., 2019; Pérez Huber et al., 2021; Powell et al., 2013; Rowe, 2008). The juxtaposed depiction of microaggressions and microaffirmations in identical STEM settings is intended to be visually engaging and instructive.

The SESSI allows individuals to reflect independently on racial microaggressions and microaffirmations, enabling them to consider their own behaviors privately and potentially reducing resistance to change. Apart from those intent on committing racial microaggressions at any available opportunity, individuals may commit them due to underdeveloped racial consciousness, unchecked stereotypes, or perhaps these combined with strained cognitive processing during tasks such as teaching or problem-solving (K. J. C. Espinoza & Rincón, 2023; Haynes & Patton, 2019; Kanter et al., 2020; Perez et al., 2023; Williams et al., 2025; Williamson, 2025). If STEM educators are shown depictions of familiar settings and learn to recognize the racial microaggressions that occur there, as well as how to enact microaffirmations, they will have cognitive scripts to access in similar situations. Further, they may be motivated to implement behavior change if they are shown how it aligns with their goals (Pope et al., 2018), such as productive interactions with students or positive teaching evaluations.

Development of the SESSI was proposed to occur through iterative phases of design-based research with prototype development and revision responsive to data collected from college STEM students and educators. Future research would test the tool’s efficacy in influencing behavior change among STEM educators, addressing a need for evidence of effective interventions for racial microaggressions (Osman et al., 2024).

The project was terminated before the end of its first year. The momentum that our team, especially the student research assistants (both graduate and undergraduate), had built up was brought to a sudden halt. Our work took place at an institution where many students would be the first in their families to graduate from a university, with hopes of breaking generations of economic struggles. We knew from the literature reviewed here and our own experiences that racial microaggressions are one more hurdle in attaining STEM degrees and careers. When the path to degree attainment is already rocky, those extra hurdles can sometimes bring you down, and you wonder whether it is worth getting up again. That is when STEM students are likely to drop out. Our plan to test an intervention for racial microaggressions dissolved, as the termination notice stated, “to protect the interests of the government.”

4. Testimonio and Other Strategies for Advancing STEM DEI Research

Having described my project on racial microaggressions, the grant’s termination, and the sociopolitical contexts surrounding my research and its termination, I turn to potential responses but not before discussing the significance of this testimonio.

Candidly, documenting the events of my project’s cancelation came with apprehension. Academics are not encouraged to detail disappointments. Failed experiments are not published, portfolios for promotion highlight accomplishments but not attempts, and hours of scrapped drafts count for nothing except lost sleep. The fact that my project ended through no fault of mine still made me hesitate to write this. I authored the grant proposal with the support of my one Co-PI, and it was a single-institution study, not a large collaborative project. As I have discussed, my research is influenced by a lifetime of both informal and formal study of educational opportunities. Therefore, when the project was halted, the blow felt personal.

With some distance, I now observe that my immediate response was to attribute this setback to an indescribable misstep on my part rather than to a structural obstacle. This is not unlike what STEM students encountering racial microaggressions experience. As the literature I reviewed suggests, Black and Hispanic/Latine students contend with underrepresentation and underachievement stereotypes in STEM. When they are confronted with microaggressions, the cognitive and emotional toll of these interactions can contribute to disengagement from STEM, especially if buffers and supports are unavailable. It can be difficult for students and STEM educators to recognize that the disengagement is often a function of structural barriers (McGee et al., 2023) rather than personal failings.

The value of this testimonio is reflected in a statement by Delgado Bernal et al. (2012): “As a process, testimoniar (to give testimony) is the act of recovering papelitos guardados—previous experiences otherwise silenced or untold—and unfolding them into a narrative that conveys personal, political, and social realities” (p. 364). I have been able to attest to an event, identifying it as personal yet tied to the sociopolitical context, with immediate and broader consequences. In line with the intent of testimonios, this testimonio has potentially offered solidarity to minoritized STEM students and researchers.

Even more difficult than detailing this defeat is conceptualizing strategic responses; playing to our strengths is perhaps the key. One response derives from the testimonio I have offered. As scholars and scientists, we understand our obligation to disseminate and critique our work. Analogously, it is necessary to document the conditions under which funded research, especially research aimed at strengthening and diversifying STEM disciplines, is upended. Moreover, as researchers, we are adept at identifying patterns, making predictions, and considering plausible outcomes. The issue is whether we are willing to harness these skills to illuminate misalignments between science and policy that threaten adverse societal effects.

Anti-DEI actions flout research evidence showing that diversifying STEM fields promotes innovation and problem-solving for an interconnected global society (Ito et al., 2025; Kamalumpundi et al., 2024; Meyer et al., 2025; National Academies of Sciences, Engineering, and Medicine, 2023). Further, decades of data identify racial disparities in STEM, employment, income, health, and other domains (e.g., Economic Policy Institute, 2025; Kaufman, 2026; Macias-Konstantopoulos et al., 2023; National Academies of Sciences, Engineering, and Medicine, 2023; Town et al., 2024; Yan et al., 2026). As scholars and scientists, we have a responsibility to caution against practices and policies that do not align with substantial research findings.

Further, NSF’s actions contradict its previous position on the value of broadening participation in STEM (American Educational Research Association, 2025). As those who apply for NSF grants are aware, one of NSF’s review criteria pertains to “broader impacts.” Previously, NSF advised applicants that “broader impacts” may pertain to “Full participation of women, persons with disabilities and underrepresented minorities in STEM” (Wilbert & Tedjasaputra, 2021). Now, National Science Foundation (2025) states that research involving “individuals or groups based on protected characteristics” is allowed “when doing so is…not focused on broadening participation in STEM on the basis of protected characteristics” (FAQs, “Can I research individuals or groups based on protected characteristics?”). National Science Foundation (2025) attributes the change to a shift in priorities established within the current “policy context” (FAQs, “Why is NSF changing its priorities?”). Let us consider that NSF was established by Congress to “promote the progress of science” and “advance the national health, prosperity, and welfare” (National Science Foundation, n.d.). When science provides overwhelming evidence of racial disparities in health, prosperity, and welfare, as well as the benefits of racial/ethnic diversity, it is evident that the current “policy context” is misaligned with science.

Arguably, with time, politics will shift, and NSF priorities will again align with science. In the meantime, we might be reassured by National Science Foundation’s (2025) recent statement: “Projects that aim to create more opportunities for more people in every region of the country to participate in science and technology remain an important part of NSF’s mission” (FAQs, “Can we still research gaps in the achievements of certain groups of people?”). However, NSF’s statement disrespects the magnitude of empirical and programmatic work (much of it funded by NSF) that addresses racial and gender inequities in STEM to build inclusive ecosystems (see, for example, Bustamante-Mora et al., 2025; Cobian et al., 2024; Z. Liu & Kringos, 2025; Mendoza-Denton et al., 2023; Pai et al., 2025). An ineffectual statement of this type, which seeks to pacify and detract from realities, amounts to race-evasiveness—sidestepping race or minimizing racial disparities (K. J. C. Espinoza & Rincón, 2023; Haynes & Patton, 2019).

When we discuss expanding participation in science in a race-evasive manner, we risk undermining efforts to diversify STEM fields or obscuring any progress. For instance, data show that Hispanic/Latine students have made gains in STEM bachelor’s degree attainment; yet, the rate of Hispanic/Latine workers in STEM occupations remains low, given their overall representation in the workforce and rising degree attainment (Yan et al., 2026). Not just the challenges but the accomplishments of Hispanic/Latine students and workers are invisible if we superficially address STEM opportunities for all people.

Additionally, continued progress in degree attainment cannot be taken for granted, especially since the 2023 U.S. Supreme Court decision (Students for Fair Admissions, Inc. v. President & Fellows of Harvard College, 2023; Students for Fair Admissions, Inc. v. University of North Carolina, 2021) ended considerations of race in college admissions. Analyses by the Urban Institute using 2018-2024 data from research universities found increases in Black students’ applications and a slight increase in their enrollment, but no increase in admissions (Cohn, 2026). The Urban Institute noted that when tracking the effects of policy changes on higher education participation, population trends and access to comparable, reliable data must be considered (Cohn, 2026). Findings by the College Board (Bloem et al., 2025) showed that following the 2023 Supreme Court decision, the share of underrepresented minority students at “more selective” private institutions declined by 2.8 percentage points. Past examination of college enrollment following statewide affirmative action bans showed dips in minority student enrollment at public universities, including several flagship institutions (H. Liu, 2022). Events of the present time demand that we continue work on educational equity. To this end, it is imperative that we ensure the collection and accessibility of data on race and ethnicity.

At this moment, the executive office has submitted its 2027 budget request to Congress (Office of Management and Budget, 2026), which proposes eliminating funding for NSF’s Directorate for Social, Behavioral, and Economic Sciences (SBE). Professional associations, including the Society for Personality and Social Psychology (2026) and the Computing Research Association (2026), are urging elected leaders to prevent the cut and members to share the impacts of their work. Thus, being vigilant and outspoken in sociopolitical contexts about the stakes of scientific research may be a new responsibility for sustaining our fields of work, not to mention achieving diversity, equity, and inclusion in STEM.