Multiple regression model
Based on data analysis, the study uses multiple linear regression equations to determine the quantitative correlation between multiple independent variables and dependent variables (see Table 8). All independent variables have significant predictive power for dependent variables. In the statistical data onto collinearity of all independent variables, the VIF value is less than 10 means that there is no collinearity between the respective variables, and the regression equation of the model is stable and reliable. The residual values of [-2, +2] can explain most of the predicted values and prove the validity of the regression equation. Among the four predictors, four variables have significant predictive power for family quality of life, followed by self-efficacy, web-based learning tools, depression-anxiety-stress and general health. The multiple correlation coefficient between the four predictors and the dependent variable family quality of life is 0.636, the coefficient of determination (R2) is 0.405, and the F value of the regression model integrity test is 46.205 (p=0.000<0.05), so the four predictions The variables can effectively explain 40.5% of the variation to family quality of life.
From the perspective of the predictive power of each variable, the most predictive factor of family quality of life is the self-efficacy independent variable, with an explanatory variation of 29%; followed by the web-based learning tool, which explains the variation 34.9%; the predictive power of the remaining two independent variables depression-anxiety-stress and general health are 37.0% and 40.5% respectively. From the perspective of standardized regression coefficients, the β values of the four predictors in the regression model are 0.336, 0.209, -0.247, 0.222, where a positive number means that its influence on family quality of life is positive, and a negative number mean its It has a negative impact on family quality of life.
The non-standardized regression equation is as follows: family life quality = 7.577 + 0.783 * self-efficacy + 0.329 * web-based learning tools-0.346 * (depression-anxiety-stress) + 0.747 * general health
The standardized regression equation is as follows: family quality of life=0.336*self-efficacy+0.209*web-based learning tool-0.247*(depression-anxiety-stress) + 0.222* general health
Table 8 Results of multiple regression analysis