| The purpose of this study was to investigate the relationship among basketball coaches' leadership behavior and team cohesion and athletes' satisfaction in selected universities in northern Taiwan. A survey questionnaire was distributed to 537 basketball athletes, 517 of which were returned for a return rate of 96.3%. There were a total of 465 survey questionnaire completed, the return rate being 89.9%.;Data was collected through the use of a self-designed Demographic Information Questionnaire and the Leadership Scale for Sports (LSS), Group Environment Questionnaire (GEQ), and Athlete Satisfaction Questionnaire (ASQ) which identify factors by which to measure leadership behavior, team cohesion, and athlete satisfaction. Results showed a strong correlation among the five leadership scales, four sub-factors of group environment, and thirteen sub-factors of athlete satisfaction. The correlation analysis showed that the training and instruction, democratic behavior, social support, and positive feedback contributed to higher levels of team cohesion and athlete satisfaction. Autocratic behavior had a negative correlation to team cohesion and athlete satisfaction. The team cohesion and athlete satisfaction had a higher correlation to each other.;The findings indicated that significant differences were found between genders in leadership behavior, and athletes' satisfaction. In addition, athletes who had the same gender as the coaches were perceived to be stronger in training and instruction, autocratic behavior, social support, positive feedback, academic support service, and external agents. There were significant differences evident in training days among leadership behavior, team cohesion, and athlete satisfaction. There were significant differences evident in training hours as respects training and instruction (f(2,462) = 4.958, p<.05), as well as social support (f(2,462) = 3.443, p<.05) of leadership behavior. Athletes perceived male coaches as exhibiting more autocratic behavior (t(463) =2.739, p<.05) and external agents (t(463) =2.574, p<.05) than female coaches. In contrast, athletes perceived female coaches as being more involved with training and instruction (t(463) = -3.647, p<.05), social support (t(463) = -1.388, p<.05), and positive feedback (t(463) = -2.647, p<.05) than male coaches. In addition, coaches that were of the same gender as the athletes were perceived to be stronger in involvement with training and instruction, autocratic behavior, social support, positive feedback, and external agents. |