Hierarchical regression model
From the results of the hierarchical regression analysis (see Table 9), it can be seen that the two independent variables of ”depression-anxiety-stress” and ”general health” in Model 1 can explain 24.8% of the variance of the dependent variable of family life quality. The F value is 45.060 (p<0.000), reaching a significant level of 0.05. It shows that the regression coefficients of the two predictors of ”depression-anxiety-stress” and ”general health” are both significant, and the standardized regression coefficient β values of these two independent variables are -0.426 (p<0.000) and 0.438 (p<0.000), respectively, both reached a significant level. Among them, the influence of depression-anxiety-stress on the family life quality is negative, and the influence of general health on the family life quality is positive. If the two independent variables of web-based learning tools and self-efficacy are further invested, the overall explanatory variation increases by 15.7% (△R2), and the F value of the significant change is equal to 35.877 (p<0.000), which means web-based learning tools and self-efficacy have significant effects on the quality of family life. The F-value of the stratified binary linear regression overall tests was 46.205 (p<0.000), reaching a significant level of 0.05. It represents that the four predictors have significant explanatory power for family life quality, and their common explanatory variation is 40.5%.
Table 9 Hierarchical linear regression results