Constructing the skills of test in Golf

• Main Authors: Jhih-Hong Chen, 陳志宏
• Other Authors: Ping-Kun Chiu
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/79676780870861311839

碩士 === 國立體育學院 === 運動技術研究所 === 95 === The purposes of the study were that Golf was separated to ten sorts of skills, and we constructed ten sorts of skills of test. The method was to run the test of ten sorts of skills of Golf with fifty-two outstanding golf athletes; on the other, we collected fifty-nine athletes’ scores by each hole which were recorded by Acer Amateur National Ranking Tournament(Winter) in 2007 that was held for four days. Criterion-related evidence quantifies a relationship between Golf skill test and the performance of games (criterion). Using Facets program to analyze data by Rasch Model, we knew that the Separation Reliability of testing items for Rasch is 0.96 and the Separation Reliability of Athlete’s estimation for Rasch is 0.91. The fitness of ten sorts of skills testing is in standard level, and this proved that there were content validity (item calibration was valid) and construct validity (measurement was valid) in this study. We used proficiency which was estimated in Golf games to be the criterion to compute the Rasch measurements of ten sorts of skills testing and the correlation of proficiency in games. We knew that the correlation is 0.65 that it reached significant level, and r-square is 0.425, which meant that ten sorts of skills testing could explain 42.5% proficiency of Golf, and proved that ten sorts of skills testing of this study had good external validity, and constructing the Norm which provided the explanation for testing original scores. At last, the conclusion of the study was that there were excellent reliability and validity studying ten sorts of skills testing by Rasch Measurement, and completed the Norm.