Regression Analysis: NBA Statistics and their Impact on Player Salary. A development of guidelines applicable to any regression analysis.
Presenter(s)
Sarah Milligan
Faculty Mentor
Dr. Una Shama and Dr. Wanchunzi Yu
Abstract or Description
When a player is a free agent, an individual who can sign with any team, one wonders what salary option will be the highest for that player. Will signing with Team A or Team B provide them with the largest salary? What factors will affect their salary the most? Does last year’s statistics have a strong impact on next year’s salary? To answer these questions, we collected data from multiple sources and complied it into one excel sheet. Then to accurately estimate and predict player’s projected NBA salary, we established a regression model using the statistical software in this project, R. Specifically, the use of a Box-Cox Transformation is considered, along with the use of t-Tests, a Brown Forsythe Test (constant variance), as well as other appropriate statistical tests and methods. We also visualized and analyzed the data in Tableau. Lastly, a comprehensive guideline for completing a regression analysis was developed. These guidelines are in depth and rigorous and can be applied to any regression analysis.