Discipline: Data Science
Subcategory: Social Sciences/Psychology/Economics
Session: 4
Room: Senate
Mariana K Duarte - University of Illinois at Urbana- Chamapign
Co-Author(s): Dr. Kelly Gaither, Texas Advanced Computing Center, Austin, Texas
The COVID-19 pandemic changed consumer behavior. Non-pharmaceutical interventions were implemented to reduce COVID-19 spread. These included mask mandates, social distancing, regulations on outdoor gatherings of over 100 people, and restrictions placed on businesses and services. While fast food service thrived during the pandemic from a gross population perspective, little is known about sub-populations disaggregated by ethnicity. Whataburger is an American regional fast food restaurant chain that began in Corpus Christi, Texas in 1950. Additionally, Texas has the second largest percentage of Hispanic/LatinX people (39.34%) in the US. The hypothesis for this research was that the non-pharmaceutical interventions that were put in place to prevent the spread of COVID-19 altered consumer behavior. To investigate this hypothesis, this research examined Hispanic and non-Hispanic consumer behavior by looking at visits to Whataburger locations in Texas using Safegraph Mobility data (SM). Machine learning algorithms: Linear Regression, K-means, and Support Vector Regression, used pre-pandemic and pandemic visitor behavior disaggregated by ethnicity to predict trends. Python was used to integrate SM data and Census data to estimate this disaggregated consumer behavior. By using census data collected in census block groups, we were able to estimate the ethnicity of visitors to Whataburger locations. The results, during the year of 2019 there was a significant increase of Latinx consumers at Whataburger locations; this increase might be influenced by the regulations implemented. The findings of this research inform consumer behavior when populations experience significant disruptive events such as the COVID-19 pandemic. Future research, will be comparing different fast food chains’ consumer trends in a variety of socioeconomic communities.
Funder Acknowledgement(s): National Science Foundation (NSF) Award #1852538 and REU Site: Cyberinfrastructure (CI) Research 4 Social Change
Faculty Advisor: Kelly Gaither, kelly@tacc.utexas.edu
Role: I did literature review of consumer trends before, during, after Covid-19. I organized the data by census block groups. Also, prepared the data to apply machine learning algorithms. Produced the graphs.