检验IRT一维性的定义分析法

Definition Analysis Method for Testing IRT's Unidimensionality

本文在仔细地考察了IRT(Item Response Theory,项目反应理论)一维性假设的含义之后,提出了一种检验其一维性假设的定义分析法。该法共分五个步骤,其主要原理是首先假定测验是一维的,然后再来验证它是否符合一维性的定义。其判定准则是,对于某一给定的IRT模型,如果在几种不同条件下所得到的“得分与能力回归曲线”吻合一致,则表明该测验是一维的。如果在穷举了几个较有代表性的IRT模型之后,仍然不能证明被试的“得分与能力回归曲线”吻合一致,则可以认为该测验不具有一维性(此时还应做其它补充分析)。此法经实际应用,发现效果良好。所以定义分析法是一种有效的方法。

This thesis gives a definition analysis method for testing IRT' s unidimensionality after studying its definition in detail. This method contains five steps. Its main principle is that we suppose the test is unidimensional first, then test whether it is unidimensional or not. Here is the testing rule, to a certain IRT's model (eg. Rasch Model), the test is unidimensional if those 'score and ability regression curves' are in good-fitness with each other. on the other hand, the test is not unidimensional if those' score and ability regression curves are not fit with each other after trying several main IRT' s model. (We had to try another method to analyze it now.)lt is proved that this method is ef fective in practice, because its testing result in practice is the same as the testing result of other effective methods (eg. factor analysis and bejar method). Thus the definition analysis method is an ef fective method for testing IRT' s unidimensionality.