A SEM with Definite Algorithm for Analysis of Social-Psychological Health of Internet Users

Structural equation model (SEM) has been widely applied in many fields in recent years. This paper builds a SEM to analyze social-psychological health of Internet users. We give the index system for the SEM and describe the relationships among the latent variables. Meanwhile we give a definite linear algorithm for SEM. Our algorithm is different from the traditional algorithms such as PLS and LISREL, and skillfully use LSE and prescription regression. In our algorithm, we first calculate the latent variables in observation system by fixed modular lengths of latent variables. Then we change the modular lengths of latent variables so that the path coefficients between any latent variable and its manifest variables to satisfy prescription conditions. With our model, the relationship between latent variables in indexes system can be analyzed, as well as the relationship between the latent and manifest variables. Consequently, the relationship between motives and behaviors of Internet users can be analyzed efficiently.