Dr. Nancy Wong, SoHE’s Kohl’s Chair in Retail Innovation and Chair of the Consumer Science department; Dr. Lydia Ashton, Assistant Professor of Consumer Science; and Dr. Cliff Robb, Associate Professor of Consumer Science, are among the researchers who have won project funding from the Center for Financial Security (CFS) at the University of Wisconsin–Madison, as part of the Retirement and Disability Research Consortium (RDRC). The center is led by Dr. J. Michael Collins, the Fetzer Family Chair in Consumer and Personal Finance at SoHE and Professor in the La Follette School of Public Affairs.
Eleven projects were awarded funding in the third year of support from the U.S. Social Security Administration (SSA) for the center’s research on the financial well-being of economically vulnerable families, older people, people with disabilities, low-wealth households, and children.
“More than ever, with Covid-19, families and individuals are struggling with both short-term financial insecurity—like just putting food on the table and keeping the lights on—and also long-term economic uncertainty, like saving for emergencies, retirement, and education,” says Dr. Ashton, whose project with Dr. Wong received renewal funding in this announcement. “Our research, and that of others with the center, addresses these concerns from a variety of angles, with an eye always to the wellbeing of the most vulnerable members of our society.”
See abstracts below for the projects by Drs. Wong, Ashton, and Robb, read the full press release about the funding announcement, and view all 11 CFS RDRC 2020-21 projects.
Applying Deep Learning to Natural Language Processing of Online Forum Conversations on SSA Programs to Improve Communication and Outreach
Text analysis of data collected from online forum conversations reveals that Social Security Disability Insurance (SSDI) applicants and recipients share concerns and confusion about the application, appeal, and continuing disability review (CDR) rules and policies. For example, many SSDI customers (e.g., veterans) do not understand how the substantial gainful activity (SGA) rule is applied. These applicants also experience significant differences across program offices and geographic regions on the application of rules and/or the interpretation of these rules (e.g., how disability, medical improvement, or SGA is defined). Preliminary results from Year 2 project suggest that confusions about how SSDI rules are interpreted and applied significantly contribute to high SSDI rejection and appeal rates. This study intends to build upon Year 2 project findings to provide insights on designing effective communication strategies to reduce confusions and aid in improving customer service experiences and welfare. Specifically, the study aims to: 1) identify the major areas of confusion about SSA rules and decision criteria using a machine-learning hybrid approach for Natual Language Processing (NLP) and text analytics, and 2) evaluate the impact of how and when SSA customers obtain such information that impact their subsequent interpretation of this information.
Assessing Vulnerability to Social Security Scams
Over the last few years, Social Security scams have become one of the most common forms of government imposter fraud (Fletcher 2019; AARP 2019). This study will develop and test an intervention to combat Social Security scams by training individuals to discriminate between scams and sincere appeals. The effectiveness of the interventions will be tested in a randomized controlled trial on a nationally representative sample of Americans. The study will employ an interactive online tool to accomplish this—a customized version of the open-source phishing security platform, Gophish, usually used to test organizational networks for security holes. The platform sends scam emails and directs participants to spoofed webpages that seek to capture their personal information. The study’s primary outcome of interest is participants’ ability to correctly distinguish scam from non-scam appeals, relative to a control group. Secondary outcome variables include level of comfort online and factual knowledge about scams.