United States Health Insurance Analysis

This project uses a simulated dataset from a machine-learning textbook whose goal was to accurately predict total annual medical costs submitted to health insurance companies in the U.S. The effects of smoking and a person’s Body Mass Index (BMI) were the focuses of the analysis but the study also included the effect of age, children being included on the insurance plan, and the region the policyholder lived in. Insurance companies calculate premiums with age, location, tobacco use, individual vs family enrollment, and the amount of coverage a person chooses to purchase. This study is important because healthcare is expensive in America and far too many people go without basic health care simply because they cannot afford it. There are bigger reasons why this is the case, but they are beyond this study. Using graphical analysis, various types of statistical analysis using the programming software R, and model building, this study will show that age, smoking, a BMI greater than 30, and having children significantly increases the total amount that people are paying in health care annually. Local and state resources for smokers and obese people will also be discussed after conclusions are drawn.