Item Parameter Estimation for a Continuous Response Model Using an EM Algorithm
Item Parameter Estimation for a Continuous Response Model Using an EM Algorithm
复制标题
使用 EM 算法进行连续响应模型的项目参数估计
DOI：
10.1177/014662169802200402
复制
发表时间：
1998
影响因子：
1.2
通讯作者：
Lingjia Zeng
中科院分区：
文献类型：
--
作者：
Tianyou Wang;Lingjia Zeng
本文用EM算法对Samejima（1973）的连续项目反应模型进行了参数估计。对该模型的潜在有用性进行了检验，包括模型的密度函数、项目参数的实际意义和统计性质。还描述了人（0）参数的期望后验和最大似然估计及其相关的标准误差。项目参数估计程序的编制和评价使用模拟数据。结果表明，该方法在估计项目和0参数方面具有较好的效果。这些估计程序应允许应用的连续响应模型的许多测量问题。
An item parameter estimation procedure is cedure was pro-developed using an EM algorithm for Samejima's (1973) continuous item response model. The potential usefulness of this model is examined, including the density function and the practical meaning of the item parameters and statistical properties of the model. The expected a posteriori and maximum likelihood estimates of the person (0) parameters and their associated standard errors are also described. The item parameter estimation pro grammed and evaluated using simulated data. The results show that the estimation procedure performs well in estimating item and 0 parameters. These estimation procedures should permit the application of the continuous response model to many measurement problems.