摘要
随着结构方程模型评价体系的推广应用,评价体系的可比性问题值得关注。本文以顾客满意度指数(CSI)为例说明结构方程模型评价体系的可比性问题,并通过模拟研究说明结构方程模型评价体系保持可比性的关键,那就是在其他条件一致的情况下,各结构方程模型中对核心变量的显变量设定一致,而各结构方程模型纳入何种潜变量以及如何设定潜变量之间关系,对评价体系可比性的影响并不显著。
The comparability among various measurement systems by structural equation model is of significance with the increasing popularity of them.However,it is an open question.This paper has done meaningful jobs on it and given a principle about the comparability.The core of comparability is the consistency of the meanings of core variable in structural equation model,which means the manifest variables specification about the core latent variable in one system should be consistent with that in another system.Meanwhile,it＇s found that the latent variables expect the core and the relationships among all latent variables have no significant effect on the comparability.
出处
《数理统计与管理》
CSSCI
北大核心
2011年第2期246-253,共8页
Journal of Applied Statistics and Management
关键词
结构方程模型
评价
可比性
顾客满意度指数
蒙特卡罗模拟
structural equation model
measurement
comparability
customer satisfaction index
Monte Carlo study
