Model–Data Fit

Empirical models play an important role in bringing order, comprehension, and manageability to complex interrelationships among variables. They enhance researchers’ abilities to develop hypotheses and provide mechanisms to speculate about multifaceted processes. In educational and related psychological research, empirical models are most often developed in order to explain latent constructs and are therefore considered to be only approximations of reality. Latent constructs are inferred based upon observable (i.e., measured) indicators or behaviors, each subject to errors in measurement. Goodness-of-fit measures, used in the context of latent construct modeling, describe how well the observed data represents the latent constructs of interest. Inferences related to latent constructs are drawn from these observable occurrences; therefore, assessing the goodness-of-fit for a model is one of the most important aspects ...