| 等级反应模型的Gibbs抽样方法 |
| |
| 引用本文： | 付志慧,郝立柱,徐宝.等级反应模型的Gibbs抽样方法[J].东北师大学报(自然科学版),2011,43(2):33-38. |
| |
| 作者姓名： | 付志慧 郝立柱 徐宝 |
| |
| 作者单位： | 沈阳师范大学数学与系统科学学院;东北师范大学应用统计教育部重点实验室;吉林大学数学研究所; |
| |
| 基金项目： | 国家自然科学基金资助项目(10971081); 辽宁省教育厅项目(L2010515); 沈阳师范大学2010年博士启动基金资助项目 |
| |
| 摘 要： | 探讨了MCMC算法在多级评分项目反应模型参数估计中的实现及其估计精度.针对等级反应模型,基于数据扩充技术,提出了一种高效灵活的Gibbs抽样方法,得到了各个参数的Markov链.随着潜在变量的引入,每个参数的满条件分布为相应参数的先验分布的截断分布.这种抽样方法适用于任何类型的先验分布,不受先验分布形式的约束.对应每个...
|
| 关 键 词： | 项目反应理论 等级反应模型 Gibbs抽样 |
Gibbs sampling method in the graded response model |
| |
| FU Zhi-hui,HAO Li-zhu,XU Bao.Gibbs sampling method in the graded response model[J].Journal of Northeast Normal University (Natural Science Edition),2011,43(2):33-38. |
| |
| Authors: | FU Zhi-hui HAO Li-zhu XU Bao |
| |
| Institution: | FU Zhi-hui1,HAO Li-zhu2,XU Bao3(1.School of Mathematics and System Science,Shenyang Normal University,Shenyang 110034,China,2.Key Laboratory for Applied Statistics of MOE,Northeast Normal University,Changchun 130024,3.Research Institute of Mathematics,Jilin University,Changchun 130012,China) |
| |
| Abstract: | This paper demonstrates the Markov Chain Monte Calro(MCMC) method,which is now widely used in parameter estimation in the item response theory(IRT) model abroad and discusses its application in the parameter estimation of polytomous IRT models.Gibbs sampling is the simplest,and the most popular MCMC method.Based on a data augmentation scheme using the Gibbs sampler,this article proposes a Bayesian procedure to estimate the Graded response model(GRM),and get the markov chains of the corresponding parameters.... |
| |
| Keywords: | item response theory graded respone model gibbs sampling |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
