Table of Contents

ABSTRACT

The National Basketball Association (NBA) attendance has been steadily increasing in recent seasons unlike other major professional sports. Also, single-game attendance data consist of multilevel determinants that are derived from individual games and nested by seasons and teams. Thus, the present study used the three-level (i.e., game, season, and team-level) hierarchical linear modeling (HLM) approach for an accurate analysis of the relationship between attendance determinants and the attendance from the 2005-06 to the 2017-18 NBA regular seasons. Among twelve game-level and thirteen season-level attendance determinants, a total of twenty predictors (e.g., home/visiting team’s quality, game uncertainty, day of game [p < .05]) significantly influenced the NBA attendance. Further, the current study statistically confirmed the recent increase in the number of spectators (i.e., Season and Season2 [p < .05]), and that the significant predictors (e.g., visiting team’s age, home team’s payroll and number of star players, home city’s population) that affected the NBA attendance during the last six seasons when the number of spectators increased were different from the variables (e.g., visiting team’s final rank, arena capacity) during the first six seasons. Moreover, this study first introduced the three-level HML approaches, which have the advantage of analyzing the multilevel data, and provides detailed statistical evidences to understand which attendance determinants significantly affected the number of spectators in the NBA and how to increase their attendance. Keywords: attendance, attendance determinants, hierarchical linear regression, National Basketball Association, NBA attendance