ABSTRACT

Researchers in the social and behavioral science areas have become increasingly interested in using structural equation models (SEM) to address substantive questions. The steps a researcher typically takes in conducting SEM applications are: (a) the researcher translates the proposed theoretical processes into a path diagram; (b) the structural equations represented by the diagram are developed; (c) data are collected and the resulting covariance matrix prepared; (d) the parameters in the structural equations are estimated using one of the available software packages; and (e) the researcher judges the adequacy of the model, via the use of goodness-of-fit indices, by comparing the difference between the sample covariance matrix and a corresponding reproduced covariance matrix based on the parameter estimates.