应用项目反应理论对瑞文测验联合型的分析  

An Item Analysis of Combined Raven's Test on the Item Response Theory

作　　者：肖玮[1] 苗丹民[1] 朱宁宁[2] 张青华[2] 

机构地区：[1]第四军医大学心理学教研室,西安710032 [2]北京师范大学心理学院,北京100875

出　　处：《心理科学》2006年第2期389-391,共3页Journal of Psychological Science

基　　金：全军指令性课题资助(编号:02L003)

摘　　要：使用BILOG-MG3.0软件,边际极大似然估计,3参数Logistic模型对354名不同能力水平的男性青年的瑞文测验联合型数据进行了分析。结果显示:大多数瑞文测验联合型的题目都适合3参数Logistic模型(有6道题不适合)。整个测验的信息函数峰值的位置在难度量表的-3到-2之间,其值为16.82。共有18道题的信息函数峰值在0.2以下。从区分度来看,72道题目的区分度均大于0.5,比较理想。难度参数显示所有题目均较低,绝大部分都在0以下,最高的只有1.01。题目的难度主要由所需的操作水平决定。伪猜测参数在0.07-0.24之间。综合分析表明瑞文测验联合型对正常青年的智力评价精度较差。To analyse the items of Combined Raven ＇ s Test （CRT）, the data of 354 male students at different ability levels were analysed with the BILOG-MG3.0 （Marginal maximum likelihood estimation, MMLE） logistic model. The results showed that all the items tallied with the three-parameter logistic model except 6 items, and the maximum information is approximately 16.82 in interval between-3 and -2. The exact location was at -2. 75, but the maximum informations of 18 items were less than 0.2. The discriminations of all items were greater than 0.5 and the pseudo guessing parameter of all items were between 0.07 and 0.24. Most difficult parameters of items were less than 0; the maximum as only 1.01. The difficulty of the items as decided by the graphic operation＇s level which the items required. The conclusion is that the CRT is not fit for the evaluation of young people.

关 键 词：瑞文测验联合型 项目分析 项目反应理论 BILOG 

分 类 号：B841[哲学宗教—基础心理学]