競技成績表現診斷評量之個案研究

The Case Study of Sport Performance Diagnostic Assessment

10.29706/GS.201208.0005

姚漢禱(Han-Dau Yau)

多層面Rasch模式 ； 成績表現 ； 效度 ； 信度 ； many-facet Rasch model ； performance ； fitting model ； validity ； reliability

大專高爾夫學刊

9期（2012 / 08 / 01）

38 - 49

繁體中文

研究目的在於診斷評量職業高爾夫運動員的競技表現。以2010及2011年度個案（Wei）在日本女子職業高爾夫賽的成績表現為研究對象。方法採用多層面Rasch模式，設定賽事、回合和球洞三個層面，估計各層面的潛能、效度、信度、適合度統計、非期望反應和偏差報告，並比較年度的差異。研究結果：一、賽事層面：樣本Rasch 個別信度為0.35，具有樣本代表性。二、回合層面：樣本Rasch個別信度為0.47，具有樣本代表性和測量效度。三、球洞層面：樣本Rasch個別信度為0.85，具有測量效度和樣本代表性。四、非期望反應和年度差異比較，皆未達顯著差異。五、偏差報告發現五桿洞優於四桿洞。結論為：利用多層面Rasch模式能夠診斷出個案競技表現的優缺點。

The purpose of study was to assess and diagnose the sport performance for professional golfer. The objects were case's performance on 2010 and 2011 Japan Ladies Professional Golf Tour in this study. The method was to use many-facet Rasch model. Three facets of analysis model were set for games, rounds and holes. The case was used to estimate the ability, validity, reliability, fit statistics, unexpected responses and bias reports, and compare both annual differences on the game and facets. The results of study: (1) games facet: the sample Rasch separation reliability was 0.35, and that had sample representativeness. (2) Rounds facet: the sample Rasch separation reliability was 0.47, andthat had sample representativeness and measurement validity. (3) Holes facet: the sample Rasch separation reliability of 0.85, which estimation reached measurement validity and sample representativeness. (4) Unexpected responses and annual differences had no significant differences. (5) The bias reports found that the par 5 was better than the par 4. The conclusion of this study was that using many-facet Rasch model could diagnose the strengths and weaknesses of sports performance in case study.