First, SEM allows a range of relations between variables to be recognized in the analysis compared to multiple regression analysis, and those relations can be recursive, or non-recursive." Thus. SEM provides the researcher with an opportunity to adopt a more holistic approach to model building. As with multiple regression and path analysis, the level of prediction and explanation can still be assessed, and hypotheses can be tested through the assessment of the significance of path coefficients. However, the judicious use of a range of measures of fit can provide the researcher with a basis for evaluating the overall model.