(1) Factor analysis was performed on all items that measure the model constructs.
(2) Based on the initial factor analysis, constructs with eigenvalues greater than one were retained.
(3) Only items with loadings of at least 0.50 were retained. Items with loadings greater than 0.50 on seven constructs were investigated thoroughly.
(4) The corrected item-total correlation was computed for each item using only the items belonging to the same
construct.
(5) Cronbach’s Alphas were computed for each construct, reliability coefficients of 0.70 or higher are considered acceptable.
(6) A CFA using SEM was performed on the final measurement model, several goodness-of-fit indexes were
used to assess the validity of the constructs