摘要
多项选择题是标准测验中最常见的题型之一,它一般包括一个正确项和多个干扰项。大多数研究中都将多项选择题处理成为"答对"、"答错"两种情况,而事实上多项选择题的干扰项中包含大量有意义的信息。本文按照题目的设计,信息的提取,量化分析的顺序,对以往多项选择题中干扰项信息的利用方法进行了回顾。其中量化分析方法又包括,干扰项的统计分析方法、多项选择题的模型化方法和使用干扰信息辅助评估的方法。文章逐一对这些方法做了总结和评价,并探讨了干扰项信息利用方法的未来发展。
Multiple - choice items are wildly used in standardized tests in psychological and educational measurement. There are one correct option and several distractors in a multiple - choice item. However, correct option means right and distractors mean wrong in most cases. In fact, different distractors in the same item contain very different information. Treating all the distractors as zero or non - master is unreasonable. In order to make distractors useful,three steps should be taken. First, develop types of multiple - choice items. Second,quantify these items. At last, choose fit statistics or models. All the methods proposed to make use of distractors in these three steps can be classified into five conditions : a. exploring the distractom by designing various items and options ; b. using data augmentation method to recode the distractom; c. constructing statistics to exploit measurement properties of distractors; d. parameterized the distrac- tors in item response models or cognitive diagnostic models; e. making distractors as external information when other methods are used. We reviewed the methods of using information of distractors in multiple - choice item in this article and discussed the problems in pres- ent existing researches.
出处
《心理学探新》
CSSCI
北大核心
2015年第3期261-265,共5页
Psychological Exploration
关键词
多项选择题
干扰项
多项选择题模型
干扰项功能差异
multiple - choice item
distractor
multiple - choice models
differential distractor functioning