インクプロットテストへの項目反応モデルの応用

抄録

In this research, a new scaling procedure was developed in order to measure personality traits (θ) to create Movement responses of organic objects (M or FM in the Rorschach test) in the Inkblot test. This procedure was based on the item response theory. In this procedure the homogeneity of the items of the used test were checked. Here, 60 items including the Rorschach test and other items from Holtzman Inkblot Technique (HIT) were used and 402 subjects were asked to create one response for each item. Then, 40 homogenious items including 8 items from the Rorschach test were selected through the principal factor method. The item parameters of the selected 40 items and the information function for the estimation of θ were then calculated. The estimates of θ were also calculated by the maximum likelihood method. A discussion on the charactristics of items and personality traits estimated on Movement took place. And the result proved that the estimate of θ (personality traits) in order to create Movement responses were not reliable at a low level.