Written by Tilde Jaques, MDI Journalism Intern
Carole Roan Gresenz is a professor at the McCourt School of Public Policy and holds the Bette Jacobs Endowed Professorship in the Department of Health Management and Policy in the School of Health. She has been working with the Massive Data Institute (MDI). As a faculty affiliate with the Massive Data Institute (MDI), Professor Roan Gresenz has been able to advance work on two research projects that combine her field of economics with cutting-edge data technology to produce valuable research to inform public policy in health, and social media and gun violence.
Prior to joining Georgetown’s faculty in 2018, Professor Roan Gresenz spent time at the RAND Corporation, a nonprofit that strives to help policymaking through research and analysis. Since joining the hilltop, she has held various positions, including Interim Dean of the School of Nursing & Health Studies from 2019-2021 and Senior Advisor for Strategic Health Initiatives in the Office of the President from 2021-2022. Now, in collaboration with MDI, she is working on a research project about the financial impact of undiagnosed Alzheimer’s Disease and another project that uses social media to fill the gap in gun violence and ownership data.
Professor Gresenz began her research on Alzheimer’s in 2015, and is continuing to work on her project with MDI Director Lisa Singh, Professor Jean Mitchell, and Dr. Dean Scott, as well as with collaborators at the Federal Reserve Bank of New York. She and Professor Singh focus specifically on the financial impact of early Alzheimer’s disease, especially before a diagnosis. As Professor Roan Gresenz explains, together, they have two goals: “one is to [get] more evidence about how potentially vulnerable Alzheimer’s patients might be prior to a diagnosis, [in order to] put in protections for people. [And the] second goal is to develop predictive algorithms using machine learning to see if we can predict who is likely to get Alzheimer’s, not just a year before a diagnosis but maybe two or three or five.”
Professor Roan Gresenz explained that her team began this research with “some anecdotal evidence and literature, which showed that one of the first functional impairments when a person starts to get the first symptoms of Alzheimer’s is limited financial capacity.” In order to transition from anecdotal to systematic evidence of the correlation between deteriorating finances and early Alzheimer’s disease, the team is merging data from Equifax, one of the largest credit agencies in the United States, with Medicare data on people’s health. Professor Roan Gresenz said that in doing this, the team is “trying to see what the financial outcomes we observe in credit data are and how those are changing before a diagnosis.”
They found that many people suffered from poor financial outcomes prior to being diagnosed with Alzheimer’s; in some cases it was because their levels of assets decreased, and in other cases it was because their debt increased. Professor Roan Gresenz explains that “this could be because they were getting worse at managing their own finances, or it could be that they were subject to scams or financial exploitation because of their cognitive impairment.”
Now, Professor Roan Gresenz explains that she is working with MDI to build “algorithms using [financial] information on people they know have Alzheimer’s, and then looking back to build other algorithms to prospectively predict for people who don’t have a diagnosis if they show signs of those financial patterns.” She explained that, while testing these algorithms, the team has to keep in mind that these algorithms may not be one size fits all; she emphasized that “when we think about these algorithms we’re seeing if we can develop one overall or if they have to be for specific subgroups.”
In addition to her research on Alzheimer’s, Professor Roan Gresenz has also been working with MDI to see if social media data can be used to track gun ownership and attitudes toward gun violence. She explained that while there is data on gun violence and general sentiment towards firearms in the U.S., “compared to other areas our data and measures in order to do research on gun policy are really limited. We don’t have a lot of national surveys, and we have not consistently collected Data on who owns guns.” One possible implication of this research is estimating the percentage of people in a state who own firearms based on an algorithm she is developing with MDI that predicts firearm ownership based on X (formerly known as Twitter) data.
To do this, Professor Roan Gresenz explained that the team wanted to “see if we can use what people are posting on twitter and predict who owns a gun based on what they’re posting.” She highlighted that this data is important because in order “to understand what we’re getting from the people posting about topics involving guns and gun violence, it can be helpful to know about who they are and whether or not they are gun owners so we can understand the representativeness of that data.”
When asked what drew her to this research topic, Professor Roan Gresenz explained that, “It’s a topic that’s kind of hard to ignore. We see the results of gun violence every day in the news, it’s a very pervasive topic. I think that whenever I’m choosing a topic for my research, I’m always interested in topics where there’s a real need for research and where there are strong [possible] policy implications.”
Overall, Professor Roan Gresenz is doing extremely interesting and important research with MDI. She describes this partnership as “very rewarding,” and mentions that “to learn the language that’s associated with a whole different field and to engage in something that’s multidisciplinary is one of the hidden joys of this collaboration.”