Trustworthy Conversational Agent Design for African Americans with Chronic Conditions during COVID-19
Kim Junhan, Jana Muhic, Sun Young Park, Lionel Robert · Deep Blue (University of Michigan) · 2021
This paper discusses preliminary findings on how to design chatbots that can increase African Americans’ trust in health information, particularly those who have experienced chronic conditions during the COVID-19 pandemic. COVID-19 has disproportionately affected the African American community in terms of severity and mortality, and scholars point towards the long-held medical mistrust among this population as a possible reason. Recent studies on the impact of conversational agents (CAs) on increasing trust in health information suggest that CAs can be effective. Through interviews and design studies with ten participants, we present four findings on how to design trustworthy CAs for our target population.