2015 年 14 巻 p. 1-10
The purpose of this study was to create an achievement evaluation standard for men’s discus throwing. Subjects were 96 male discus thrower. The six trials of the discus throwing movement were recorded by camera, and the best record was used for analysis. Delphi method with 8 experts was implemented for characteristic causal analysis. Item characteristic and test characteristic were analyzed by IRT of 2PLM. Binary data (success=1, failure=0) was collected following the achievement decision criterion. The discus throwing was classified in the main of 4 motor phase (preparation, first turn, second turn, release).The discus throwing movement was configured 11 sub-movements and 34 items. A difficulty gap between the first turn and the release turn was large, there was a lot of items than other phases. Therefore, it is guessed that the first turn phase and the release phase are main phase and important for discus throwing. Unidimensionality, fitness to 2PLM and invariance of parameters were confirmed. The test reliability coefficient displayed a maximum value of 0.97 at an ability of -0.1. Ability range with the reliability coefficient of over 0.8 was between -0.2 to 2.0. It is means that the achievement evaluation standard of the discus throwing movement skill was high test precision. The correlation of the ability and the throwing record was 0.93 (p <0.05). The validity of this achievement evaluation standard was confirmed.