Representation and inclusion among members and affiliates of the Society for Epidemiologic Research: findings from the 2021 diversity and inclusion survey

Representation and inclusion are stated priorities for many scientific and professional organizations, including the Society for Epidemiologic Research (SER), which was founded in 1967 with the intention of bringing together epidemiologists across career stages and specialties.1,2 Representation and inclusion are necessary for fostering safe and equitable educational and professional environments, recruiting future generations of researchers and practitioners, and addressing critical public health questions. However, there has been persistent underrepresentation and systemic exclusion of marginalized groups in the sciences, including epidemiology,2,-12 which are symptoms of interpersonal and structural racism, classism, sexism, ableism, heteronormativity, religious-based discrimination, and other dimensions of marginalization.13,14 To advance SER’s goals, the Diversity and Inclusion Committee has sought to characterize representation and inclusion among SER members and affiliates through surveys conducted in 2018 and 2021. In this study, we assessed trends in representation within SER, made comparisons with relevant benchmarks, and discussed barriers to inclusion. In the 2018 baseline survey, many groups were underrepresented relative to the US population, particularly transgender individuals, Black/African American and Hispanic/Latinx people, and first-generation college students.11 Moreover, women and people with certain racial/ethnic and religious identities were less likely to participate in SER activities or to report feeling welcomed. This letter provides primary findings from an updated assessment of representation and inclusion among SER members and affiliates and situates the experiences of SER members in broader literatures on diversity and inclusion. The complete report, including a more detailed discussion of recommendations informed in part by survey respondents, is available on the SER website15 and eScholarship Repository.16 We invited SER members and affiliates (former members and individuals who had participated in SER activities) to participate in an online survey between May and September 2021. The survey was adapted from the version administered in 201811 and comprised sociodemographic questions across several domains, including individual, educational, career, and institutional, as well as perceptions of inclusion in SER. To compare representation in SER with the field of epidemiology more broadly, we obtained demographic data on race/ethnicity from the Association of Schools and Programs of Public Health (ASPPH) for all faculty and US citizen students attending US-based institutions. To compare representation among survey respondents born or residing in the United States with US population diversity, we obtained nationally-representative sociodemographic data from the 2020 census, with the addition of data on representation of Middle Eastern/North African individuals from the Los Angeles Times17 and sexual/gender minorities from the Williams Institute at UCLA.18 To examine academic institutional diversity, we used data from the 2021-2022 US News and World Report (USNWR) Global University Rankings, which appraised 1750 institutions globally and ranked them based on peer assessments and bibliometric indicators.19 We also identified which universities met the criteria for Minority Serving Institutions (MSI), including Historically Black Colleges and Universities (HBCUs) and Hispanic-Serving Institutions (HSIs). We (D.J.X.G. and V.P.P.) translated the present article to Spanish (Appendix S1), the second most commonly spoken language among respondents. We convened the full authorship team to discuss our positionality and wrote a collective statement (Appendix S2), as the positionality of researchers can influence the choice of topic, epistemological framing, ontology, methodological approaches, connection to participants, and communication of findings.20,21 The survey instrument is included as Appendix S3. There were 1148 respondents to the 2021 survey; 79.3% were current SER members, 15.1% were former members, and 5.6% had no membership history but participated in SER-sponsored events (Table 1). The 2018 survey had 631 respondents. Between 2018 and 2021, there were increases in the proportion of survey respondents who were Hispanic/Latinx (8.9% in 2018, 10.0% in 2021) and Asian/South Asian (17.1%, 18.5%). The proportion of Black/African American (10.2%, 7.7%) and Native Hawaiian/Pacific Islander (1.9%, 0.6%) respondents decreased, although the number of respondents from these groups increased. Among SER affiliates residing in the United States, compared with the general US population, there was underrepresentation of people who identified as Hispanic/Latinx (8.4% in SER, 18.5% in the United States) and Black/African American (8.6%, 13.4%) (Table S1). While there was little change in the proportion of female or male respondents, there was greater representation from nonbinary (0.2% in 2018, 1.0% in 2021) and transgender individuals (0.2%, 1.3%) (Table 1). The proportion of bisexual (4.3% in 2018, 6.4% in 2021) and queer (0.6%, 9.0%) individuals increased. There was also growth in representation of first-generation college students (12.8%, 20.9%). The most frequent countries of birth or residence for SER affiliates were the United States, Canada, Brazil, China, and India. The proportion of respondents who speak a language other than English at home also increased (17.6%, 27.6%). Concerning disability, 3.8% of respondents indicated a physical disability in the 2018 survey, and in the 2021 survey, 2.0% indicated that they require disability-related accommodations. Descriptive statistics for respondents to the 2021 and 2018 diversity and inclusion surveys,a as well as the 2021 and 2018 membership surveys (representing all Society for Epidemiologic Research members) aData on sexual orientation, language use, and several specific racial/ethnic identity were not collected in all membership surveys. bRace/ethnicity categories are not mutually exclusive; survey respondents could select all that applied. cWe provide more detailed information on country of birth or country of residence in Table S3. There was an increase in respondents’ perceptions of diversity of ideas at SER, but a decrease in the proportion who felt that SER was inclusive or diverse. In both surveys, individuals from a religious group felt less welcomed; given the sensitivity of this issue, we elected not to disclose which group. Compared with women, men had more positive perceptions of attempts to feel welcomed, feel comfortable sharing opinions, that members accept diversity, that early-career researchers were included, that members communicate respectfully, and that there were opportunities for collaboration, with the highest sense among White men. White individuals also were more likely to report positive changes concerning the extent they felt welcomed and more positive perceptions of the diversity of institutions represented at SER than other respondents. Among the 815 survey respondents with academic affiliations, 54.0% were affiliated with institutions in the highest-ranked quartile of the USNWR Global University Ranking distribution (higher-ranked), and 8.2% were affiliated with institutions in the lowest-ranked quartile (lower-ranked). In both the 2018 and 2021 surveys, over a quarter of respondents were affiliated with 8 higher-ranked US academic institutions (Table S2). Participants affiliated with lower-ranked institutions were more likely to be first-generation college students (35.2% vs. 16.4% in higher-ranked) (Table S3). Lower-ranked institutions were more likely to be publicly funded, in non-urban settings, or outside the United States (Table S4). Fewer than 1% of respondents were affiliated with Historically Black Colleges and Universities (HBCUs). Fourteen percent of respondents were at Hispanic-Serving Institutions, of whom 10.3% were Hispanic/Latinx. Among epidemiology students in the ASPPH dataset, there was higher representation of Hispanic/Latinx (11.3% in ASPPH, 8.4% in SER) and Black/African American (12.4%, 8.6%) scholars compared with US-residing SER affiliates (Table S5). In the 2021 survey, we observed increased representation of some groups but persistent underrepresentation and exclusion of certain marginalized groups. Specifically, there was less representation of Black/African American and Hispanic/Latinx people among SER members compared with these groups’ representation among epidemiology students and the general US population. The proportion of American Indian, Alaska Native, Native Hawaiian, and Pacific Islander individuals in SER was similar to that of the US population. There was low representation of Minority Serving Institution–affiliated scholars and underrepresentation of people with disabilities.22,23 Institutional diversity did not improve substantially between the 2018 and 2021 surveys, with disproportionately high representation of higher-ranked US institutions. This disparity has also been noted for presenters at SER’s Annual Meeting,2 and is likely related to the wide financial disparity between the most well-resourced private universities and public and non-US universities.24 The field of public health, and epidemiology in particular, has grown substantially in recent decades, with some increases in representation from persistently marginalized groups. Between 1992 and 2012, graduates in public health programs have been more diverse than the general population of undergraduate students, with increasing representation from some racially marginalized groups.3,-5 However, among epidemiology faculty, there was no change in representation of racially marginalized groups between 2010 and 2020, with persistent underrepresentation of American Indian, Alaska Native, Native Hawaiian, Pacific Islander, Black/African American, and Hispanic/Latinx individuals.6 Indeed, in the ASPPH dataset, both for students and faculty, there was decreasing representation of Black/African American individuals in increasingly senior roles. With respect to gender, prior work has also found that women are less likely to be last authors on peer-reviewed epidemiology papers, to be cited by colleagues, and to serve in leadership roles in epidemiology journals compared with men.25 An assessment of presenters at SER’s Annual Meeting from 2015-2017 found underrepresentation of women in invited symposia and overrepresentation of presenters from private institutions.2 More work is needed for gender minorities. Persistent underrepresentation and exclusion of marginalized groups are not unique to epidemiology and occur in other biomedical disciplines. A 2018 study of US medical school faculty found a statistically significant increase in underrepresentation of Black and Hispanic individuals across nearly all medical disciplines between 1990 and 2016, using a novel statistical technique.26 Among registered US nurses, despite slight increases in racial diversity from 2008 to 2018, there have been persistent issues in recruiting and retaining nursing professionals who reflect their diverse patient populations.8,9 A study of diversity among vascular surgeons found that women, Black/African American, and Hispanic/Latinx providers were persistently underrepresented at all career stages, with notably less diversity in senior positions.10 The Society for Pediatric Anesthesia had disproportionately low representation in more senior positions of racially marginalized people, women, gender minorities, and sexual minorities.27 The exclusion of racially and socioeconomically marginalized people in higher education and the biomedical sciences has been reported in other countries represented in SER.28,-34 Exclusion of people with marginalized identities has broad impacts, perpetuating harms to individuals and communities and impairing the quality of the scientific enterprise at large. People from marginalized groups have higher rates of scientific innovation, but the contributions of these individuals are less likely to be recognized, and consequently these scholars are less likely to remain in scientific careers or to obtain research-oriented professorships.35 Misconceptions of the extent of marginalization may exacerbate exclusion of marginalized groups.36 Constraints in our assessment include restriction of ASPPH student data to US citizens attending US institutions. Data specific to public health training outside the United States could elucidate further disparities and barriers. The USNWR Global University Rankings are subject to bias, although we addressed this by comparing institutions by quartile. The survey response rate was lower among men compared with people with other genders, as with other web-based surveys.37 We used the questions on requested accommodations as proxy indicators of disability status; specific questions about disability status and experiences of interpersonal and structural ableism in professional settings would provide more complete and actionable data. Given shifting demographics among the diverse populations epidemiologists serve, future researchers should consider applying statistical measures such as the S-score to assess underrepresentation among trainees and professionals.26 The S-score is a numerical expression of the probability of an observed proportion of representation of, for example, certain racial/ethnic and gender identity groups in a professional setting based on proportion of those groups in the overall population.26 Issues of representation and inclusion are cross-cutting, and SER leadership and members have already undertaken initiatives to address persistent interpersonal and structural barriers. These initiatives include engagement with faculty and trainees at institutions under-represented in SER through the SERvisits program24 and outreach to underrepresented non-US institutions, such as the 2023 SER Mid-Year Meeting at the National Institute of Public Health in Cuernavaca, México. We found that while there is diverse representation across many dimensions of identity, underrepresentation and exclusion persist for certain groups. It may be that the benefits of diversity and inclusion efforts have not yet been realized, that these efforts are ineffective, and that additional barriers or biases remain unaddressed.24,38 Furthermore, as evidenced by the underrepresentation and exclusion of marginalized groups in other biomedical disciplines,2,3,26 broader structural issues influence the experiences of current and prospective SER members. Additional careful work is needed to understand experiences of marginalization and exclusion within SER and the professional spaces we inhabit. Longitudinal mixed-methods studies, incorporating quantitative survey data and qualitative free-response or interview data, would improve understanding and better inform future initiatives. More work is needed to understand experiences of people with intersecting marginalized identities, those facing religious-based discrimination, individuals with disabilities, and the diverse experiences of exclusion within racial/ethnic groups.24,39 Collaboration with trained diversity and inclusion scholars can enhance the quality and impact of this work. We would like to thank Helena Archer, Hassan Bokhari, Julia Mewha, and Christopher Navarrete for assistance with data analysis, as well as Suzanne Bevan for administrative support. We are grateful to SER affiliates who participated in the survey. Members of the Society for Epidemiologic Research Diversity and Inclusion Committee: Nadia N. Abuelezam, Amira Aker, Sarah Andrea, Onyebuchi A. Arah, Obefemi Babalola, Lisa Bates, Charles Branas, Monique Brown, Yvette Cozier, Timothy Crawford, Geetanjali Datta, Elizabeth A. DeVilbiss, Mary V. Díaz Santana, Luther-King Fasehun, Lindsay Fernandez-Rhodes, David Fink, Cara L Frankenfeld, Stephen Gilman, David J. X. González, Yueh-Ying Han, Juan Hincapie-Castillo, Dayna A. Johnson, Marynia Kolak, Christina Ludema, Meghan D. Morris, Stefania Papatheodeorou, Marcia Pescador Jimenez, Victor Puac-Polanco, Abhismitha Ramesh, Domonique Reed, Jamaica Robinson, Brooke Staley, Alvin Thomas, Caroline Thompson, Jennifer Weuve, and Mahsa Yazdy (July 2020 to June 2023). Supplementary data is available at American Journal of Epidemiology online. D.J.X.G. was supported by the Ford Foundation Postdoctoral Fellowship and the University of California President’s Postdoctoral Fellowship. E.A.D. was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD. D.M.R. was supported by the National Institute of Mental Health of the National Institutes of Health under Award Number R36MH132489. B.S.S was supported by the Ford Foundation and the Genetic Epidemiology of Heart, Lung, and Blood Traits Training Grant (T32HL129982). Support of L.F.R. to present preliminary results at the 2022 Annual Meeting of the Society for Epidemiologic Research was provided by the Pennsylvania State University, Department of Biobehavioral Health. D.S.F. received support from the National Institute on Drug Abuse (K99DA055724; PI: D.F.). H.S.A.M. received support from the National Heart, Lung, and Blood Institute (K99HL161479; PI: H.S.A.M.). The authors declare no conflicts of interest. The views expressed in this article are those of the authors. Primary data that we collected through the representation and inclusion survey data are not available, due to confidentiality. Data from the Association of Schools and Programs of Public Health (ASPPH) are available to member programs and schools, and to others upon request through the ASPPH Data Center ([email protected]). Additional data sources used in this analysis are publicly available, including sociodemographic data from the US Census Bureau, UCLA Williams Institute, and the Los Angeles Times.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.