As shown in , the results show no problems in multicollinearity
because the coefficients of tolerance represent over 0.1 as well as the
coefficients of MAX-VIF also show below 10.0. Therefore, there are no
problems in multicollinearity.
2. Test of Reliability and Validity
There are three preconditions for a multivariate analysis such as
normality, homoscedasticity, and linearity. The first is tested using the
Kolmogorov-Smirnov test. When the hypothesis is rejected, a variable
shows normality (z = 1.031, p = 0.238). The second is tested by comparing
the variance of an independent variable with the variance of a dependent
variable. As a result of the analysis, the relationships between the variables
show identical transformation and this means that there are no problems in
homoscedasticity. The third is concerned with causality between the
variables. As a result of the analysis, there is no problem in linearity
(140.229 in F value, 0.393 in R2, 11.842 in t value and 0.000 in p value).
Therefore, the preconditions for a multivariate analysis are achieved in the
results.
Next is internal consistency concerned with the items and this is
analyzed by inter-item correlations (criterion: over 0.3) and item-total
correlations (criterion: over 0.5). Commitment has six items (the former:
0.669 to 0.829, the latter: 0.795 to 0.868, and Cronbach’s alpha: 0.948).