关键词:  贝叶斯, 频率论, 潜变量建模, Mplus

Abstract:  Bayesian statistical methods is one of the two statistic schools. Recently, Bayesian statistical methods are becoming ever more popular in social and behavioral research. However, domestic psychological and behavioral scholars are not familiar with it. We provide an untechnical introduction about Bayesian statistics, particularly in Bayesian method used in latent variable modeling. Specifically, first we compared the difference between Bayesian methods and Frequentist methods in several basic concepts. Thereafter, the Bayes’ Theorem and its analysis procedure were introduced. Finally, to illustrate Bayesian methods in latent variable modeling (i.e., confirmatory factor analysis), a concrete sample was presented. In the end of the paper, we briefly discussed the future direction of Bayesian methods in psychology research.

Key words:  Bayesian, Frequentist methods, latent variable modeling, Mplus