迫选式人格测验的传统计分与IRT计分模型

发布时间：2019-01-18 16:34

【摘要】：迫选测验的传统计分方式会产生自模式数据,不能进行传统的信效度检验、因素分析和方差分析等。近年来研究者提出了一些基于项目反应理论的计分模型,如瑟斯顿IRT模型和MUPP模型等,它们可以规避自模式数据的弊端。瑟斯顿IRT模型方便进行参数估计,模型定义灵活;而MUPP模型的拓展性较差,参数估计的方法有待提高。另一方面,已有研究者基于MUPP模型开发了一些抗作假的迫选测验,而瑟斯顿IRT模型距离这种应用还比较远。此外,两个模型的适用性和有效性都有待更多的实证研究来检验。
[Abstract]:The traditional scoring method of the forced selection test results in the self-pattern data, which cannot carry out the traditional reliability test, the factor analysis, the variance analysis and so on. In recent years, the researchers have put forward some scoring models based on the theory of project response, such as the Seston IRT model and the MUPP model, which can avoid the disadvantages of the self-model data. The Thesson IRT model is convenient for parameter estimation, and the model definition is flexible; and the extension of the MUPP model is poor, and the method of parameter estimation is to be improved. On the other hand, some anti-fraud forced-selection tests have been developed by the researchers based on the MUPP model, while the Thesson IRT model is far from the application. In addition, both the applicability and the effectiveness of the two models are to be tested by more empirical research.
【作者单位】：北京师范大学心理学院应用实验心理北京市重点实验室;中国基础教育质量评价与提升协同创新中心;
【基金】：教育部人文社会科学研究青年基金项目(11YJC190016) 国家自然科学基金项目(31100759) 北京市与中央在京高校共建项目(019-105812)资助
【分类号】：B848

【共引文献】