The GME ,PLS ,MLE and GSCA estimation methods for the structural equation models

The study of the customer satisfaction (CS) in the health care service, usually called patient satisfaction, is typically developed by using the structural equation models (SEM). Traditional approaches in estimating SEM parameters have raised two important issues: the first concerns with the Maximum Likelihood (ML) approach developed by Jöreskog (1973) and is often called “hard modeling”, because of the distribution assumptions and large sample; the second one concerns with a free distribution approach and small sample, namely, Partial Least Squares (PLS) by Wold (1982) and also called “soft modeling”. Recently, two new free distribution approaches are introduced for SEM parameters estimation, namely Generalized Maximum Entropy (GME) by Al Nasser (2003) and Generalized Structured Component Analysis (GSCA), developed by Hwang and Takane (2004). The aim of this paper it to show the GME method for the structural equation models, showing how this method can be considered a valid alternative to the PLS (Al Nasser 2003, Ciavolino et al, 2006). Moreover a new comparative study is developed with the GSCA and ML methods. The next sections start with a brief introduction to the estimation methods, followed by a motivating example on health care service, and comparative comments on the empirical results and considerations by a methodological point of view.