Turning Anger Into Action: Minority Students Analyze COVID Data on Racial Disparities

As the coronavirus swept into Detroit this spring, Wayne State University junior Skye Taylor noticed something striking. On social media, many of her fellow Black classmates who live or grew up in the city were “posting about death, like, ‘Oh, I lost this family member to COVID-19,’” said Taylor.

The picture was different in Beverly Hills, a mostly white suburb 20 miles away. “People I went to high school with aren’t posting anything like that,” Taylor said. “They’re doing well, their family is doing OK. And even the ones whose family members have caught it, they’re still alive.”

How do COVID-19 infection rates and outcomes differ between these ZIP codes? she wondered. How do their hospitals and other resources compare? This summer, as part of an eight-week research collaborative developed by San Francisco researchers and funded by the National Institutes of Health, Taylor will look at that question and other effects of the pandemic. She’s one of 70 participants from backgrounds underrepresented in science who are learning basic coding and data analysis methods to explore disparity issues.

Data to address racial discrepancies in care and outcomes has been spotty during the pandemic, and it isn’t available for most of these students’ communities, which disproportionately bear the brunt of the virus. The participants are “asking questions from a perspective that we desperately need, because their voices aren’t really there in the scientific community,” said Alison Gammie, who directs the division of training, workforce development and diversity at the National Institute of General Medical Sciences.

Scientists from Black, Hispanic, Native American and other minority backgrounds have long been underrepresented in biomedicine. By some measures, efforts to diversify the field have made progress: The number of these minorities who earned life science doctoral degrees rose more than ninefold from 1980 to 2013. But this increase in Ph.D.s has not moved the needle at the faculty level.

Instead, the number of minority assistant professors in these fields has dipped in recent years, from 347 in 2005 to 341 in 2013. And some of those who have entered public health endure racial aggression and marginalization in the workplace — or, after years in a toxic environment, quietly leave.

“We really need to focus on making sure people are supported and find academic and research jobs sufficiently desirable that they choose to stay,” said Gammie. “There have been improvements, but we still have a long way to go.”

In 2014, the NIH launched the Building Infrastructure Leading to Diversity initiative. It offers grants to 10 undergraduate campuses that partner with scores of other institutions researching how to get poor and minority students to pursue biomedical careers.

Students in the program receive stipends and typically spend summers working in research labs. But when COVID-19 hit, many labs and their experiments shut down. “People were like, what do we do? How do we do that remotely?” said biologist Leticia Márquez-Magaña, who heads the initiative’s team at San Francisco State University.

She and University of California-San Francisco epidemiologist Kala Mehta sketched out a plan for students to work remotely with bioinformatics, population health and epidemiology researchers to collect and analyze COVID-19 data for marginalized populations.

Gammie encouraged the Bay Area team to expand the summer opportunity to participants across the nation. From June 22 to Aug. 13, students spend two to three hours online four days a week in small groups led by master’s-level mentors. They learn basic bioinformatics — computational methods for analyzing biological and population health data — and R, a common statistical programming language, to collect and analyze data from public data sets. “I think of basic bioinformatics and R coding as an empowerment tool,” said Mehta. “They’re going to become change agents in their communities, fighting back with data.”

Bench science often takes years, whereas data crunching to solve problems offers a sense of immediacy, said Niquo Ceberio, who recently earned a master’s in biology at SFSU and leads the team of mentors. “There was this sort of limitlessness about it that really appealed to me,” she said.

Raymundo Aragonez, a University of Texas-El Paso biology major participating in the summer program, sees data analysis as a way to address confusion in the Hispanic community — including some of his family members who think the pandemic “is all a hoax.” Dismayed by misleading YouTube videos and rampant misinformation shared on social media, Aragonez, who aims to be the first in his family to finish college, said he hopes to gain skills to “understand the data and how infections are actually happening, so I can explain it to my family.”

He hopes to explore whether COVID-19 infection rates differ among people living in El Paso, those living in the Mexican city of Juárez, and those who frequently cross the border between the cities — like many of his friends and classmates.

Willow Weibel, an SFSU psychology major, is studying how COVID-19 restrictions affect the mental health of former foster youth and other young adults with traumatic backgrounds. Weibel spent much of her childhood in foster care before getting adopted into a Southern California family at age 17. “I’ve grown to really care about what other people go through in the system,” she said.

Mental health is a common thread in the research questions proposed by several students in Weibel’s group, including Skye Taylor, who is majoring in psychology with a minor in public health. While curious about disparities in Detroit-area COVID-19 outcomes, she also wants to examine how mental health issues affect COVID-19 susceptibility — “especially in the Black community because mental health isn’t really talked about,” she said.

Having the chance to explore their own research questions is unusual for undergraduates, and particularly meaningful to students of color. “It feels like science is something that’s been done to us or on us,” said Ceberio, who is Black and Latina, and grew up in Los Angeles, Miami, and Las Vegas before moving to the Bay Area. “This experience allows them to do research that they feel is relevant based on the way they’re viewing the world. I’m trying to get them to trust their instincts.”

Trainees from underrepresented groups will more likely stay in biomedicine if they feel they are giving back to their communities or doing something with a tangible purpose, said Gammie. This summer, participants “have an opportunity to engage in science that does both,” she said. “Our hope is that this will inspire students to go on to be independent scientists.”

Designing a biomedical coding program focused on inclusivity and agency

To the Editor — Last year, scientists across the world participated in the 10 June protests and used hashtags such as #ShutDownAcademia, #ShutDownSTEM and #Strike4BlackLife, which called for non-Black academics to take action towards eradicating anti-Black racism in academia, including the disciplines of science, technology, engineering and math (STEM). This resulted in an outpouring of stories by Black academics detailing the racist experiences that they have endured. Their stories underlie an urgent need for transforming academic structures to be more equitable.

To make a positive change towards this goal, last year we created the Virtual BUILD Research Collaboratory (VBRC), an online summer program focused on diversity and inclusion. Summer programs are crucial for students as they are a critical step towards building scientific careers. The VBRC was developed out of the SF BUILD (Building Infrastructure Leading to Diversity), a six-year-old program for undergraduate students from diverse backgrounds in biomedical research, typically held in person. Given the disruptions caused by the pandemic, our team had to transform the regular in-person offering into an eight-week online summer coding program, which took place from 22 June–13 August 2020. As we were able to pivot rapidly to a new solution, we offered it to nine sister National Institutes of Health (NIH) BUILD sites across the country. The objective of the virtual eight-week summer coding program was to teach students to use the statistical programming language and software R/RStudio to perform research related to health disparities due to COVID-19.

Purposeful racial representation and comprehensive inclusivity were at the center of the VBRC. The program had a total of 65 students, who were on average 23.7 years old (standard deviation of 5.25), 73% cisgender female, 37% Black, 39% Latinx, 14% non-Black/Latinx, and 10% Asian or other. Most of the students had no prior experience with coding, and many were majoring in biomedical sciences at minority-serving institutions across the United States. Additionally, students participated in scholarly engagement discussions with 15 scientists researching COVID-19 or critical race theory, most of whom were from historically underrepresented communities. When engaging with students, the faculty not only explained their current work in response to COVID-19, but also described their scientific journey. This scholarly engagement fostered a collaborative learning environment that empowered students to consider their experiences and culture as valuable sources of knowledge to advance scientific inquiry.

An intentional non-hierarchical teaching environment focused on the students’ interests was established. The VBRC program combined large group sessions (with all 65 students) with smaller breakout groups (5–7 students) led by near peer mentors (NPMs), all of them occurring online. The relationships that NPMs built with students were vital to the success of our program. NPMs were all graduate students and Black, Indigenous and people of color (BIPOC) (67% Latinx, 11% Black, 11% Asian American, and 11% Indian American), to which most of the students of the program could relate. NPMs also had learned to code within the last 2–4 years: they could easily relate to students facing challenges while learning to code and could use this experience to better guide students in this process. NPMs allowed us to stay away from the pitfalls of traditional teacher–student power dynamics. We felt that this model was important in creating an environment where students felt comfortable in asking questions in real-time, allowing them to get to a place of understanding more quickly.

VBRC emphasized real learning, motivation and presence, not grades. All of the students were living the reality of how COVID-19 was affecting communities of color1, so picking this timely, relevant and relatable theme was very important. The focus was on engaging with real problems in real-time, where students used publicly available data to understand how health disparities were affecting their communities. The final deliverable of the course was a presentation on COVID-19 and race/ethnicity: 54 team presentations were delivered with a great deal of enthusiasm and included a message to communities of color about the results on a myriad of topics (wealth disparities, prisons, Immigration and Customs Enforcement detention facilities, race/ethnicity, mental health and their relationship to COVID-19). For example, one message included the importance of COVID-19 testing for communities of color along with the availability of testing sites in the local communities.

Lastly, VBRC faculty practiced cultural humility, with an openness to learning from students about which teaching methods did not work and adapting to students’ learning styles. Adaptations varied depending on what the students, not faculty or NPMs, felt worked best. An example of this was the original curriculum, which used open-source didactic lectures to teach the R programming language. Students provided feedback that the course did not sufficiently teach the basics needed to be confident using R, and the open-source lectures were eliminated. Perhaps in a less inclusive format, students would be expected to adapt to the preset curriculum even if it did not meet the students’ needs. In VBRC, we listened to what was working and what was not working and adapted. The VBRC coordinators and NPMs then collaborated to make worksheets that introduced the basics (variables, scripts, comments and logical operators) in a step-by-step exploratory approach. Students responded positively to the worksheets, so they were adopted going forward. Extensive reading sets were eliminated and instead the instructor relayed the theoretical background using relatable examples, in one instance using data about a popular artist’s ticket sales.

Some limitations do deserve comment. Most of the students (80%) reported that they wished to continue the program. As this was a summer-only program, that was not possible. A more formal way to create an affirming space for coding could be a sustained effort in the future that carries the ideas from the summer program throughout the academic year. Also, due to the NPM structure, a limiting aspect of the program is recruiting a large number of NPMs that students can relate to, as there may be few at any given university. Recruiting NPMs from multiple sites to reach more students, like the crowdfunding solution proposed by Michael Johnson and colleagues2, may be a potential way to expand the program.

Models like VBRC could be developed and tested further as a potential solution to morph computer science programs, bioinformatics programs and biomedical programs into becoming more inclusive and affirming of BIPOC students, their interests and their lived experiences. This is certainly not the only program designed as an inclusive summer program; for example, programs such as the Big Data Summer ProgramMeyerhoffMARC and RISE also focus on diversity and inclusion. As VBRC was held online, it allows for easy scaling, which can be beneficial. With multiple, combined and sustained efforts like these, we can begin to tear down the structures of racism in education, with the goal of ultimately changing the face of science.