College or School
Faculty Research Advisor
Second Faculty Research Advisor
Due to COVID-19 causing nationwide lockdowns, foot traffic in the United States changed drastically. The restaurant industry, seafood especially, was and still is impacted by COVID-19, lockdowns, and other factors. Foot traffic data is useful for companies to have but it is expensive, not available in real time, and hard to get. During times of uncertainty, it is helpful to know how many people are going to be visiting a business. The goal of this project is to find other variables that are free, available in real time, and easily accessible that can predict foot traffic for seafood restaurants in the United States. To do this, many different types of models were fit in R and compared using AIC and mean squared error. The model with the lower mean square error was a generalized linear model containing the covariates: Google search hits for 'seafood takeout', unemployment rate, and the average number of covid cases per week. The data set in this project contains weekly data from 1/5/2019-8/14/2020. The next step in this project will be to test the selected model on the most recent foot traffic data once it becomes available.
Beaulieu, Isabel P., "Finding what covariates most accurately predict foot traffic in seafood restaurants" (2022). Undergraduate Research Conference (URC) Student Presentations. 526.