The Application Of Structural Equation Model And Its Latent Variable Interaction Effect Analysis In Psychsocial-epidemiology

In psychsocial-epidemiology, researchers may often encounter some variables that exist objectively but could not be measured directly, such as cognition, capability, intention and social economic status, and so on. These variables are called latent variables in structural equation model. The inter-relations among these latent variables could not be tackled with traditional statistical analysis methods, and the relations involved in such latent variables and observed variables could not be done too. Nevertheless, as a combination of multi-regression and factor analysis, structural equation model is a multi- variables statistical analysis mean. One of its advantages is that it can be applied to analyze the practical problem with latent variables. In other hand, the status of psychsocial-factors is relative in psychsocial-epidemiology. Psychsocial-factors may rise as independent variables or dependent variables. This special situation that a variable may be emerged as a cause and a result simultaneously in the same context resulted in researchers would confront many dependent variables in one practical study. When researchers want to understand the action mechanism among variables clearly, it is insufficient only to study the direct effects or total effects involved in variables. It should be focused on to analyze the indirect effects among variables. So, researchers should take change of direct effects, indirect effects and total effects of variables. Hereby, in psychsocial-epidemiology practical study, it is an important facet to analyze interaction effects between variables. Interaction effects may be existed in observed variables, and also may be existed in latent variables. Classical statistical analysis methods could neither deal with practices with latent variables, nor do interaction effects in latent variables. At the same time, it is true that the classical statistical analysis methods could not be applied to study the indirect effects among variables. We introduced structural equation model and its latent variable interaction effect analysis into psychsocial-epidemiology practical study based on that the strengthening path coefficient theory in structural equation model could tackle indirect effects of variables, and that nonrecursive model in structural equation model has the capability to deal with the special difficult modeling problem in psychsocial-epidemiology practice, i.e., there...