Using Data From Countless Tweets, Project Including UMD Researcher Will Create National Map of Bias
A UMD researcher is studying whether racism in an area, measured through the content of local tweets, can result in health problems at birth.
The unfiltered public forum of Twitter may present a new tool to help understand the relationship between racial hostility and health in a geographic area.
Supported by a new $3.3 million grant from the National Institutes of Health, a University of Maryland researcher will help analyze the content of tweets to characterize the racial climate in regions across the United States. Those results will illuminate the team’s study of disparities in birth outcomes, including preterm births and low birth weight, both of which their earlier research indicated can be affected by area-level racial bias.
“Racial discrimination is often studied as an individual-level experience, but we know that it is pervasive and has social-environmental components,” said Quynh Nguyen, an assistant professor of epidemiology and biostatistics and a co-investigator on the study. “In the past, we have linked living in communities with higher racial hostility to a higher risk of adverse birth outcomes and cardiovascular disease.”
The project led by Thu Nguyen (Quynh Nguyen’s twin sister) at the University of California, San Francisco brings together experts in epidemiology, health disparities, machine learning, social media data, biostatistics and community-engaged research fields. They will use a range of online and social media data coupled with machine learning models to create two measures of area-level racial bias, and then employ analytics to see if the bias correlates to birth outcomes.
The team also aims to identify any protective factors for adverse birth outcomes, and follow changes in area-level racial bias, particularly when local or national events of racial significance occur.
Quynh has made other connections between health and what comes out our in 280-word Twitter blasts over the years:
A study she led that was published in the Journal of Racial and Ethnic Health Disparities based on tens of millions of race-related tweets, she and colleagues showed that people living in states with the highest number of racist tweets—including Mississippi, Louisiana and Michigan—had elevated rates of high blood pressure, diabetes, obesity, stroke and heart attack.
An earlier study, meanwhile, showed how the estimated 22% of American adults who use Twitter talk about food can correlate to area-level measures of cardiovascular health like obesity and mortality.
Maryland Today is produced by the Office of Marketing and Communications for the University of Maryland community on weekdays during the academic year, except for university holidays.
Faculty, staff and students receive the daily Maryland Today e-newsletter. To be added to the subscription list, sign up here:Subscribe