ABSTRACT
Meanwhile, to ensure that the instrument has reasonable construct validity, confirmatory factor analyses (CFA) were used. Following Campbell and Fiske’s (1959) criteria, we tested for construct validity and convergent and discriminant validity. The results reveal that the correlations are all higher than zero and large enough to proceed with discriminant validity. Furthermore, discriminant validity, proposed by Aldawani and Palvai (2002) and Campbell and Fiske (1959), was conducted by counting the number of times an item correlates higher with items in other factors than with items in its own factor. The results confirm adequate discriminant validity. In conclusion, the dimensions used in this study demonstrate both convergent and discriminant validity. Tables 7.4 to 7.11 present the outcome of the second-order CFA, including convergent and discriminant validity for each dimension.
