criteria (i.e. none included 1.0), providing evidence of discriminant validity.
In a second assessment of discriminant validity, we followed the advice of
Fornell and Larcker (1981) by assessing whether the variance extracted
estimates associated with each possible construct pair are greater than the
square of the correlation between the two constructs. Once again, evidence
of discriminant validity is provided as all variance extracted estimates
exceed the appropriate squared factor correlation. Overall, then, we
concluded that the scales had adequate measurement properties, and we
proceeded to test the hypothesized structural relationships
criteria (i.e. none included 1.0), providing evidence of discriminant validity.
In a second assessment of discriminant validity, we followed the advice of
Fornell and Larcker (1981) by assessing whether the variance extracted
estimates associated with each possible construct pair are greater than the
square of the correlation between the two constructs. Once again, evidence
of discriminant validity is provided as all variance extracted estimates
exceed the appropriate squared factor correlation. Overall, then, we
concluded that the scales had adequate measurement properties, and we
proceeded to test the hypothesized structural relationships
การแปล กรุณารอสักครู่..