Demographics
Overview
Underrepresented Populations
Race & Ethnicity
Disability | Health & Service Differences
Disadvantaged Background
Sex & Gender
Sexual Orientation
Dive Deeper
Underrepresented Populations
Underrepresented populations have historically been excluded from participating in biomedical science.
This exclusion may be intentional or unintentional, but may impact data quality, how we perceive, understand, or report on public and population health, the equity of programs, funding, public policy, and much more.
Racial and ethnic groups underrepresented in biomedical research are defined as those that are
Black or African American
Hispanic or Latino
American Indian or Alaska Native
Native Hawaiian or other Pacific Islander
Individuals with disabilities are defined as those with a physical or mental impairment that substantially limits one or more major life activities
Individuals from disadvantaged backgrounds are defined as those who meet two or more of the following criteria:
Experienced past or present homelessness
Previously or presently in the foster care system
Eligible for the Federal Free and Reduced Lunch Program for 2+ years
First generation college students
Eligible for Federal Pell grants
Received WIC as a parent or child
Grew up in a rural area: either HRSA or HPSA locations (address/zip-related eligibility)
NOT-OD-20-031: Notice of NIH’s Interest in Diversity. (2019, November 17). National Institutes of Health. https://grants.nih.gov/grants/guide/notice-files/NOT-OD-20-031.html
Practice Recommendations for Training Programs
from Mekinda and colleagues (in review)
from Mekinda and colleagues (in review)
Consider the Notice of NIH’s Interest in Diversity (NOT-OD-20-031) as a starting point for developing demographic data practices. Expand data collection as scientific and pragmatic needs dictate.
Increase granularity of demographic data to improve data accuracy and understanding of important differences between demographic subgroups.
Include open-response items for trainees to describe the demographic identities they choose, in their own words, to increase data accuracy and support emerging research. Include prompts for pronouns and sexual and gender minorities (Morrison, Dinno, and Salmon, 2021).
Allow trainees to self-report demographic identities to increase data accuracy; Adopt strategies for data validation as needed to limit underreporting.
Minimize power dynamics through sensitivity to the circumstances under which demographic data are collected.
Mekinda, M.A., Chaudhary, S., Vanderford, N.L., Burns White, K., Kennedy, L., and L.K. Marriott. (2022). Approaches for measuring inclusive demographics across Youth Enjoy Science cancer research training programs. In peer review with the Journal of STEM Outreach.
Considerations for specific demographic groups
Not everyone receives the same access to or quality of healthcare, which results in race, ethnicity, language, and disability minority groups experiencing avoidable health inequities.
Greater resolution of data collection practices can help:
Identify health inequities in subpopulations
Guide development of culturally specific and accessible services
Guide equitable allocation of resources to address inequities
Source: McGee, M.G. (2020). Race, ethnicity, language and disability (REALD) implementation guide. Portland, Oregon: Oregon Health Authority, Equity and Inclusion Division. https://sharedsystems.dhsoha.state.or.us/DHSForms/Served/le7721a.pdf
Race & Ethnicity
In an effort to improve demographic data collection standards, Oregon Health Authority’s Race, Ethnicity, Language, and Disability (REALD) offers a validated tool for collecting demographic information on race, ethnicity, language, and disability.
Notable race & ethnicity questions from REALD:
Question 1: an open-ended question encourages respondents to identify in the way they choose without the limitations of predetermined categories.
Question 2: respondents are asked to select their racial and ethnic identities from 34 options. Another write-in option is available here to self-describe race and/or ethnicity. This helps to track emergent populations.
Considerations:
Pros: REALD is an inclusive instrument; REALD answer options can be easily mapped to NIH categories for grant reporting (START, 2022)
Balance: May take longer to complete; privacy concerns with small sample sizes
Disability (Health & Service Differences)
Disabilities may be seen or unseen by others and may seriously impact an individual's quality of life. Individuals with disabilities may experience discrimination and challenges living in a society not built for them, which impacts health and function. Disability data collection helps to identify health and service differences to eliminate preventable social and health inequities.
Prompt for Participants:
“Your answers will help us find health and service differences among people with and without functional difficulties. Your answers are confidential.”
2-10 questions about hearing, vision, movement, communication, daily living, cognition, and mental health.
Considerations:
Pros: Inclusive tool; REALD answer options could be upcoded to NIH disability definitions; provides important information about how to provide accommodations to participants.
Balance: Takes longer to answer
Disadvantaged Background
Disadvantaged background often describes socioeconomic or environmental conditions that impact access to education or training environments. Disadvantaged background definitions have evolved over time. NIH issued definitions in 2018 and changed them in 2019 (NOT-OD-20-031). NIH measures relate to socioeconomics and class.
Experienced past or present homelessness
Previously or presently in the foster care system
Eligible for the Federal Free and Reduced Lunch Program for 2+ years
First generation college students*
Eligible for Federal Pell grants
Received WIC as a parent or child
Grew up in a rural area defined by either HRSA* or HPSA* locations (address/zip-related eligibility)
*Eligibility was underreported by students when verified (Marriott et al., 2022).
Marriott, L.K., Shugerman, S.R., Chavez, A., Crocker Daniel, L., Martinez, A., Zebroski, D.J., Mishalanie, S., Zell, A., Dest, A., Pozhidayeva, D., Wenzel, E.S., Omotoy, H.L., Druker, B.J., Shannon, J. (2022, in press). Knight Scholars Program: A tiered three-year mentored training program for urban and rural high school high school students increases interest and self-efficacy in interprofessional cancer research. Accepted to the Journal of STEM Outreach.Considerations:
Pros: Definitions and general conditions are described. Suggested wording via START.
Balance: No clearly defined approach. Students under-report their eligibility when verified (Marriott et al., 2022). Some backgrounds are missing (e.g. immigrants, refugees)
Consider adding: “We realize we may have not captured everything about your background or experience. If you would like to say more, please feel free to share your story” [open-field] (Marr, 2021).
Sex & Gender
Conflating sex and gender in research practices excludes entire populations whose physical characteristics and/or identities do not fit within the constraints of male/female. The erasure of sex and gender minority groups makes it impossible to provide services and accurate data for these populations. Treating sex and gender as independent variables allows researchers to collect more accurate data.
Hart, C. G., Saperstein, A., Magliozzi, D., & Westbrook, L.. (2019). Gender and Health: Beyond Binary Categorical Measurement. Journal of Health and Social Behavior, 60(1), 101–118. https://doi.org/10.1177/0022146519825749
Recommended Approach: Open-ended prompts are most supportive of gender and sex diversities and can be qualitatively coded (see START, 2021; Learn Qualitative Methods). If coding an open-ended prompt is not possible (i.e., large studies), we suggest using the two question approach from Morrison, Dinno, & Salmon (2021).
What is your gender identity? (select all that apply):
Feminine/Woman/Girl
Masculine/Man/Boy
Nonbinary/Genderqueer/Third Gender
Agender/Non-gender
Questioning
Prefer to self-describe:________
Prefer not to answer
Are you transgender?
Yes
No
Questioning
Prefer not to answer
Sexual Orientation
Sexual orientation exists on a spectrum and is independent from gender identity. The two are often conflated in research studies. Measuring sexual orientation separately from gender and sex provides more accurate representation of populations.
Suen, L. W., et al., (2020). What Sexual and Gender Minority People Want Researchers to Know About Sexual Orientation and Gender Identity Questions: A Qualitative Study. Archives of Sexual Behavior, 49(7), 2301–2318. https://doi.org/10.1007/s10508-020-01810-y
Area for Growth: Sexual orientation is not included among NIH definitions of underrepresentation. In the program sample we studied, most training programs did not ask about sexual orientation (Mekinda et al., in review). Research shows these populations face significant health disparities and discrimination; further inclusion is warranted.
Mekinda, M.A., Chaudhary, S., Vanderford, N.L., Burns White, K., Kennedy, L., and L.K. Marriott. (2022). Approaches for measuring inclusive demographics across Youth Enjoy Science cancer research training programs. In peer review with the Journal of STEM Outreach.
Recommendations: Include options to select more than one item from the list; include options to self-describe to support the fluidity and complexity of people’s identities; provide a table containing sexual orientation definitions (example found on START website)
Diamant et al.(2000) http://dx.doi.org/10.1001/archfami.9.10.1043, Katz-Wise et al. (2016), https://doi.org/10.1080/00224499.2014.1003028, & Suen et al. (2020), https://doi.org/10.1007/s10508-020-01810-y.
The survey question below is intended to support accurate identification and inclusive data collection for sexual and gender minorities (SGMs).
Regardless of your sexual experience, how do you identify your sexual orientation? (select ALL that apply)
- Straight or heterosexual
- Gay
- Lesbian
- Bisexual
- Pansexual
- Asexual
- Queer
- Fluid
- Questioning/unsure
- Prefer to self-describe: ______
- I don't know what this question is asking
- Prefer not to say
Language
While we didn't talk about language in the presentation, communication barriers associated with limited English proficiency can lead to health inequities, including quality of care and adverse hospital events that result in physical harm.
Populations with native languages other than English have historically been underrepresented in STEM and biomedical sciences.
Preferred language and English proficiency questions can be a predictor of an individual’s ability to access services and programs. The REALD Implementation Policy offers functional language questions for service-based systems and demographic language questions for non-service-based systems.
Try this
"We realize we may have not captured everything about your demographics. If you would like to say more about communities you represent, please feel free to share." (open field; phrasing from Dr. Mollie Marr to be inclusive of demographic and identity categories not represented (F30MH118762).]
Big Picture Takeaways
Research definitions of diversity evolve, reflecting the process of science. There are still populations that are being excluded, marginalized, or improperly aggregated. Look for them. Include them. Advance science.
Flexibility in demographic data collection is needed (e.g., due to time burden, participate age, funder requirements, program reach, etc.). Consider ways to be more inclusive in your data collection approach.
Be kind to yourself and others as we all learn to do better. This is the process of science.