Abstract by Micaela Johnson
Obesity Rates in Georgia: A Case Study on Reducing Americaâ€™s Obesity Epidemic
We develop a process to identify counties in Georgia with exceptionally low obesity rates despite their demographics and culture. A careful analysis of these counties may help identify certain traits or programs that help reduce obesity. We fit a Bayesian logit regression with a spatial random effect, which we call “cultural effect.” Our model accounts for the following demographics: inactivity, crime rate, % black, % latino, % female, median income, birth and death rate, migration rates, average education, rural-urban, and economic typology. Even though the parameters have a very low effective sample size, the draws of the adaptive MCMC algorithm converge. We identified two counties, Irwin and Crawford; we were unable to find any community programs or distinguishable demographics that could explain their lower obesity rate.