Page 391 - ISC PROCEEDINGS 21.4
P. 391

explained = 67.70%. For the three moderating variables: KMO = 0.910, Bartlett's test χ² =
                  1512.449 (p < 0.001), three factors extracted, cumulative variance explained = 71.48%.
                  For the dependent variable: KMO = 0.790, Bartlett's test χ² = 275.859 (p < 0.001), one
                  factor extracted, variance explained = 59.92%. All factor loadings exceeded 0.50,
                  confirming convergent validity and factorial structure.
                        4.3. Descriptive statistics and correlation analysis
                                                  Table 3. Descriptive statistics
                     Variable                           Mean                 SD
                     Sustainable Purchase (PGP)         3.971                0.715
                     Price Perception (P)               4.165                0.715
                     Green Knowledge (GK)               3.441                0.908
                     Perceived Quality (PQ)             4.025                0.730
                     Social Influence (SI)              3.805                0.616
                     Green Marketing (GM)               3.910                0.808
                     Influencer (INFLU)                 3.612                0.846
                     FOMO                               3.525                0.969
                     Financial Literacy (FL)            3.830                0.885
                        As shown in Table 3, Pearson correlation analysis revealed that all five independent
                  variables and three moderating variables were significantly correlated with the
                  dependent variable (p < 0.01), with the exception of Social Influence (r = −0.002, p =
                  0.980). Green Marketing exhibited the strongest bivariate correlation with sustainable
                  purchasing behavior (r = 0.713), followed by Perceived Quality (r = 0.641), Price
                  Perception (r = 0.608), Green Knowledge (r = 0.572), and Influencer endorsement (r =
                  0.558).
                        4.4. Regression analysis
                        The hierarchical moderated regression model in Table 4 demonstrated satisfactory
                  overall fit: R² = 0.634, adjusted R² = 0.606, F(11, 141) = 22.249, p < 0.001. The Durbin-
                  Watson statistic (1.988) fell within the acceptable range, confirming the absence of
                  autocorrelation. All VIF values were below 3.0, indicating the absence of multicollinearity.
                  The Kolmogorov-Smirnov test confirmed residual normality (Sig. = 0.200 > 0.05).
                                      Table 4. Regression results (Dependent variable: PGP)
                    Variable                    B           Std. Error    β          t         Sig.
                    Constant                    −1.202      1.208         —          −0.995    0.321
                    Price (P)                   0.202       0.055         0.201      3.652     0.000
                    Green Knowledge (GK)        0.128       0.042         0.163      3.048     0.003
                    Perceived Quality (PQ)      0.259       0.083         0.264      3.120     0.001
                    Social Influence (SI)       0.130       0.041         0.122      3.170     0.000
                    Green Marketing (GM)        0.507       0.160         0.573      3.168     0.002
                    Influencer (INFLU)          0.242       0.081         0.250      2.987     0.004
                    FOMO                        0.265       0.080         0.266      3.313     0.000
                    Financial Literacy (FL)     0.194       0.056         0.196      3.464     0.000
                    PQ × INFLU                  0.155       0.051         0.155      3.039     0.000
                    GM × FOMO                   0.059       0.015         0.620      3.933     0.000
                    SI × FL                     −0.041      0.072         −0.226     −0.569    0.570


                        Notably, Social Influence (SI) showed a near-zero bivariate correlation with PGP (r =
                  −0.002, p = 0.980), but emerged as a statistically significant predictor in the multivariate

                                                                                                      390
   386   387   388   389   390   391   392   393   394   395   396