摘要
目的:旨在介绍贝叶斯结构方程模型的特点及其使用方法。方法:首先讨论了贝叶斯结构方程模型的优势,然后以运动员训练状态检测量表(32×7)的测评数据,分别采用最大似然估计和贝叶斯估计进行二阶验证性因素分析。结果:纳入交叉载荷和残差相关等小方差先验信息的贝叶斯估计模型拟合良好,而采用最大似然估计的模型拟合不理想。分析造成上述差异的原因,并总结贝叶斯结构方程模型的优势和不足。
Purpose: To introduce the characteristics of Bayesian structural equation model and its usage. Methods:Firstly,the advantages of the Bayesian structural equation modeling were discussed in this paper,and then,a second-order confirmatory factor analysis( CFA) was carried out by using maximum likelihood estimation and Bayesian estimation with the data from the Athlete Training State Test Scale( 32 × 7). Results: The Bayesian estimation model incorporating small variance prior information such as cross-load and residual correlation fitted well,but the model fitting using maximum likelihood estimation was not ideal. This study analyzed the reasons for the above differences and summarized the advantages and disadvantages of Bayesian structural equation modeling.
作者
晏宁
李英
李玉磊
郭璐
毛志雄
YAN Ning;LI Ying;LI Yu-lei;GUO Lu;MAO Zhi-Xiong(Psychology College,Beijing Sport University,Beijing 100084,China;Psychological Services Center,Beijing Union University,Beijing 100101,China;Beijing Sino-Freneh Experimental School,Beijing 100095,China)
出处
《北京体育大学学报》
CSSCI
北大核心
2018年第9期75-82,共8页
Journal of Beijing Sport University
关键词
贝叶斯方法
结构方程模型
验证性因素分析
交叉载荷
残差相关
Bayesian analysis
structure equation model
confirmatory factor analysis
cross-loading
residual correlation