The measurement model analysis followed Koufteros (1999) paradigm for SEM and Koufteros et al. (2009) paradigm for second-order models. Koufteros et al. (2009) recommended the evaluation of four alternative measurement models when second-order constructs exist. The first model considers all indicators tied to one first-order construct; the second model hypothesizes each first-order construct, with its own in- dicators, uncorrelated (orthogonal); the third model is similar to the second model, but evaluates correlated first-order con- structs; and the fourth model considers complete second-order constructs. The goodness of fit indices for adequacy were calculated for each measurement model to compare them and determine the best model. Table 7 shows the results.