多维题组反应模型:多维随机系数多项Logistic模型的应用拓展

发布时间：2018-01-24 03:12

  本文关键词： 多维目标能力 多维题组 two-tier模型 MRCMLM 估计精度　出处：《心理学报》2017年12期 　论文类型：期刊论文

【摘要】：本文将多维随机系数多项Logistic模型(MRCMLM)拓展应用到多维题组领域,得到适用于多维目标能力和多维题组效应的多维题组反应模型(MTRM),该模型具有高度灵活性和适用性。本文通过两个模拟研究和一个应用研究探究MTRM参数估计精度和模型适用性,以及与two-tier模型的差异,结果发现:(1)能力维度间相关和项目评分等级是影响模型参数估计的重要因素;(2)MTRM对项目参数估计准确性和稳定性高于two-tier模型,对题组效应大小估计更为准确。(3)MTRM在考虑项目内多维题组情况下模型拟合度更高,为测验分析提供了更广泛的模型结构选择,具有显著的应用价值。
[Abstract]:In this paper, the Multidimensional Random coefficient Multiterm Logistic Model (MRCMLM) is extended to the field of multidimensional problem group. A multi-dimensional problem group response model suitable for multi-dimensional goal ability and multi-dimensional problem group effect is obtained. This model has high flexibility and applicability. In this paper, two simulation studies and an application study are conducted to explore the accuracy of MTRM parameter estimation and the applicability of the model, as well as the differences between the model and the two-tier model. The results showed that the correlation between ability dimensions and item score were the important factors affecting the estimation of model parameters. The accuracy and stability of MTRM in estimating project parameters are higher than that in two-tier model. It is more accurate to estimate the effect of item group. The model fit degree is higher when considering the multi-dimension problem group, which provides a more extensive choice of model structure for test analysis. It has remarkable application value.
【作者单位】： 北京师范大学中国基础教育质量监测协同创新中心;北京师范大学心理学院;
【基金】：全国教育科学“十二五”规划2013年度教育部青年课题(EBA130370)资助
【分类号】：B841
【正文快照】： 1研究背景Wang,1999;MaraisAndrich,2008;Sireci,Thissen,Wainer,1991;WainerWang,2000;Yen,1993)、导致1.1文献综述等值误差和收缩误差(Lee,Kolen,Frisbie,Ankenmann,传统测验中,如果项目之间存在关联,被试对2001;Li,Bolt,Fu,2006)、导致项目区分度参数的不同项目的反应，

本文编号：1459012