Exploratory factor analysis (EFA) was conducted to draw
the factors’ structure, with 21 observatory variables, to
confirm such variables’ reliability and to eliminate the
inadequate observatory variables. EFA was conducted
prior to confirmatory factor analysis (CFA). After EFA, the
deduced factors were defined as patent variables [16]. EFA
with oblique rotation (direct oblique) and factor extraction
with maximum-likelihood estimation were performed [15].
Six factors were extracted from EFA, which could explain
77.58 % of the overall variation and which were statistically
significant with respect to the p value and RMSEA.
Cronbach’s alpha was also over 0.6, implying the credibility
of the selected variables (Tables 3, 4).