National Basketball AssociationMultiple Linear RegressionModel SelectionMulticollinearityInfluential PointOutliers
It is of great interest to identify the factors that influence the salaries of National Basketball Association (NBA) players. This study examines the 2017-2018 wages of 100 NBA players which are randomly selected by the SAS software based on their career performance variables using a multiple linear regression. There are 28 explanatory variables which include age, 3-point field goals per game and free throws per game. The multiple regression analysis is conducted to determine the explanatory variables which are helpful in predicting the salaries of NBA players. Five methods for model selection are used, these include forward selection, backward elimination, stepwise selection, adjusted R-square selection method and C(p) method. All five methods demonstrated similar results. Results indicated that variables such as games started, field goals per game, total rebounds per game, personal fouls per game, also the terms of contract used, have a significant correlation with salary.