Structural equation model and comparisons of model fit indexes

Structural equation model is a comprehensive statistical analysis method which is used to test causal relationships between unobserved variables. This method has become increasingly popular in Turkey, as well as all over the world in recent years. This method is particularly used to evaluate relationships between variables in psychology, sociology, education, science, economics and marketing. In this study, among nearly 40 goodness of fit indexes used to determine appropriateness of models in structural equation model, variability of 14 goodness of fit index was examined according to sample size, parameter estimation method, number of factors in the model, in an effort to determine which goodness of fit index is more advantageous in what sample size. Variability of goodness of fit indexes was examined in different sample sizes in the structural equation models derived from different parameter estimation methods by both simulation studies and analyses using real data. Among 14 indexes examined in the study, only RMSEA, RMSR, and IFI indexes were found advantageous for researchers.