基于CTT、GT、IRT的评分者信度研究——以某届奥运会女子跳水决赛为例  

The Study of Raters' Reliability Based on CTT,GT and IRT:Taking Women's Diving Final in Olympic Games for Example

作　　者：钟晓玲[1] 康春花[1] 陈婧[1] 

机构地区：[1]浙江师范大学教师教育学院

出　　处：《考试研究》2013年第5期41-52,共12页Examinations Research

基　　金：浙江省哲学社会科学规划基金(项目编号:10CJY15YBB);浙江省教育厅课题(Y2010117786);浙江省教师教育基地2011招标项目的阶段性研究成果

摘　　要：本文以某届国际奥林匹克运动会女子跳水决赛为例,综合应用CTT、GT和IRT三大测量理论进行评分者信度分析,从不同角度揭示评分者之间和评分者内部的差异情况。结果表明:CTT的评分者信度分别为0.981和078;GT的概化系数和可靠性指数分别为0.8279和0.8271,比赛中所采用的7名评委分别对选手在5轮上的跳水表现进行评定的决策是比较适宜的决策;在IRT中,相对而言,评委5在7名评委中最为严厉,评委2最为宽松,但评委之间在宽严程度上的差异不显著,评委1和评委4在自身一致性上存在问题,不同评委在评定不同选手、不同难度系数动作和不同轮数上存在偏差,但未达到显著性水平。基于本文的分析,可以了解三种评分者信度分析方法的特点及各自优势,为评分者培训和提高评分信度提供有用信息。This paper takes the international Olympic Games Women＇s Diving Final as an example. It applies the three different kinds of measurement theories which are CTF, GT and IRT to analyze raters＇ reliability contributes to reveal inter-rater differences and intra-rater differences from different angles. Results show that the raters＇ reliability of CTT is 0.981 and 078 ; Generalizability coefficient and index of dependability in GT are 0. 8279 and 0. 8271. In the diving event, the seven raters＇ ratings for five rounds are comparatively appropriate. In the IRT, relatively speaking, rater 5 grades the severest, rater 2 grades the most lenient in the seven judges. But there is no significant difference in the severity among the judges. Rater 1 and rater 4 have some problems on the consistency of evaluation. The deviation of different raters＇ grading exists in different players, movement of different difficulty, different rounds. But the deviations do not reach the significant level. We can know the characteristics and advantages for the three method of raters＇ reliability analysis based on the analysis of the case. It also provides helpful information for raters training and improving raters＇ reliability.

关 键 词：评分者信度 经典测量理论 概化理论 多面RASCH模型 跳水决赛 

分 类 号：G424.74[文化科学—课程与教学论]