类别变量的结构方程模型及其应用

摘要

目的 探讨类别变量的结构方程在实际问题中的应用.方法 运用LISREL软件中的类别变量分析技术,对452个关于生活方式与健康的调查样本进行了结构方程模型分析.结果 量表中反映生活方式的8个条目一致地聚合为生活方式这一潜在因子,此因子与抑郁和焦虑呈现显著负相关关系,表明生活方式越健康产生精神障碍的几率就越低.结论 观测变量的类型是结构方程建模必须予以重视的关键要素之一.%Objective To explore the techniques of using structural equation model with categorical observed variables in substantive research. Methods Data were collected from 452 subjects using lifestyle and a 4-item measure of depression and anxiety questionnaire. Structural e-quation model with categorical observed variables was applied to the data a-nalysis. Results All 8 lifestyle-related indicators pointed to one factor called lifestyle, which significantly correlated to depression and anxiety. Conclusion Variable types deserve more attention in the applications of structural equation model.