条目反应理论简述

A Brief Introduction to Item Response Theory

条目反应理论认为,每个测验条目有其自身固有的特征,比如难度、区分度等,这些特征与样本无关,也与其它条目无关。不同的潜在特质,适用于不同特性的条目来测量,这将提高测量精度和测验效率。条目的得分概率与潜在特质呈特定的回归关系,这些关系通常称为各种数学模型。本文还简要讨论了运用条目反应理论的注意事项。

Item Response Theory takes the hypothesis that every test item has its own features such as difficulty, discrimination and others, which are not related to the sample and neither to other items. A given level of latent trait could be measured by specially featured items, so that the measurement accuracy and test efficacy will be increased. There is a certain regression between the score probability of an item and the latent trait, and the regressions are usually represented as mathematic models. This paper provided some brief discussions about the application of item response theory also.