Page 86 - ISC PROCEEDINGS 21.4

10.12). This result indicates that, although the model has been improved, autocorrelation
remains; therefore, the authors continue to remove variables with high autocorrelation to
enhance the stability and reliability of the regression estimates.
* The time-series regression model produces the following results: The linear
regression results of the model with the dependent variable SD and the independent
variables DE, CKH, CCN, LD are presented in the table, showing that the model has high
explanatory power. The R² value = 0.9928 and the adjusted R² = 0.9839 indicate that
approximately 99.3% of the variation in the Sustainable Development Index (SD) is
explained by the independent variables in the model.
Table 4: Regression model results
SD Coef. Std. Err. t t P>t
DE 0.970931 0.262694 3.7 0.006
CKH 0.0000506 1.91E-05 2.65 0.046
CCN -0.010707 0.02638 -0.41 0.706
LD -0.758742 0.209582 -3.62 0.022
_cons -0.698375 3.230999 -0.22 0.839
Source: Author
* Tests for model issues, including multicollinearity, heteroskedasticity, and
autocorrelation, yield the following findings:
The multicollinearity test, after removing variables with high collinearity, shows that
the VIF value is 3.68 (< 10), indicating that the model does not suffer from
multicollinearity.
Table 5: Multicollinearity test results

Source: Author
- The heteroskedasticity test shows that Prob > chi2 = 0.7678 (> 0.05), suggesting
that there is no heteroskedasticity problem in the model.
The results indicate that three factors significantly affect green economic
development, including digital economy development, expenditure on science and
technology, and human capital. Based on the estimation results obtained using Stata 20,
the following conclusions can be drawn.
Table 6. Results of the heteroskedasticity test

Source: Author

85