Predictive Models of the Amount of Physical Activity Across the United States

Abstract

How physical activity affects individuals' health across the United States is an important topic for the well-being of the country. The purpose of this study is to examine how certain factors influence the amount of physical activity performed by adults in each state. Using statistical analysis techniques, such as linear regression and model selection, a model of physical activity, including several covariates, is constructed. These covariates come from various categories, such as, geographic, financial and economic, residents’ living habits, and residents’ health. The final model, model (1), consists of four covariates which are obesity, refund, uninsured and poverty. This model can be used to predict how the percent of physically active adults changes as these four covariates change.