Factor Analysis and Latent Variables Models

It is known that a latent variable is a numerical variable having Z-scores, which is a variable with zero mean and unit variance, in a theoretical sense. Having a small sample size, the data analysis can be done using EViews. The data analysis should be conducted in two stages. The first data analysis is to generate the set of endogenous and exogenous latent variables, where each of the latent variables is generated based on a set of simple and observable attributes or indicators. A second-level latent variable should be generated based on a set of first-level latent variables, which have been applied by Hamsal and Ary Suta in their dissertations. In this chapter, the theoretical concept of the factor analysis is not presented in detail, but how to generate a defined latent variable is covered. Controlled Vocabulary Terms factor analysis; latent class model