Several items were eliminated from the models because of
poor fit and Table 1 provides the final measurement model fit
indexes for each of the five measurement models. As can be
seen, the fit indexes obtained were good to excellent. As
additional tests for discriminant validity, a series of two-factor
CFA models was conducted for each possible pair of measures.
For each model, the correlation between constructs was set at
unity, and then freed. In all cases, the χ2 decreases were
significant.
Appendix A provides a listing of all items used for model
testing. For completeness, we also report the results of a ‘full
measurement model’ (see Table 1) in which all the remaining
scale items were entered simultaneously into a CFA with the
Several items were eliminated from the models because ofpoor fit and Table 1 provides the final measurement model fitindexes for each of the five measurement models. As can beseen, the fit indexes obtained were good to excellent. Asadditional tests for discriminant validity, a series of two-factorCFA models was conducted for each possible pair of measures.For each model, the correlation between constructs was set atunity, and then freed. In all cases, the χ2 decreases weresignificant.Appendix A provides a listing of all items used for modeltesting. For completeness, we also report the results of a ‘fullmeasurement model’ (see Table 1) in which all the remainingscale items were entered simultaneously into a CFA with the
การแปล กรุณารอสักครู่..