PROBABILITY AND STATISTICS IN ITEM ANALYSIS AND CLASSIFICATION PROBLEMS Statistical Decision Theory Approach to Item Selection for Dichotomous Test and Criterion Variables

摘要

In the present paper, a nonparametric model is formulated in the framework of statistical decision theory, in the case that both the test variables and the criterion variable are dichotomous. In this case, the unknown criterion variable can be regarded as a classificatory variable whose value specifies to which of two possible classes an individual belongs