Correlations used the Pearson Product Moment approach for continuous variables and Point-Biserial analyses for nominal data. SEM was selected as the major analysis because this approach evaluates values for the key outcomes through generating factor scores that reflect the characteristics of each population and minimizes error variance [42]. The SEM was carried out with the maximum likelihood estimation for analysis of the variance/covariance matrix of AMOS [43]. The final SEM was also run using Mplus as a cross-check and to use the attributes of both programs [44]. The stability of parameter estimation and of the structural equation model results were evaluated via AMOS bootstrapping procedures. This was repeated 1000 times. Goodness of fit for the measurement model and SEM included a determination of the chi square (Π2); the Π2 to degrees of freedom ratio (good fit = a 2-to-3 ratio) [45]; the CFI, or Comparative Fit Index (good fit is indicated by scores > .90) [46, 47]; the NNFI or Non-Normal Fit Index with good fit indicated by values approaching 1.0 [48]; RMSEA or the Root Mean Square Error Approximation
