Table 5.
Standardized direct and indirect pathways (n = 865).
| Mechanism | Effect | BootSE | 95% CI | Contribution rate (%) | |
|---|---|---|---|---|---|
| LLCI | ULCI | ||||
| Total effect | 0.453 | 0.029 | 0.397 | 0.509 | 100 |
| Direct effect | 0.216 | 0.024 | 0.169 | 0.262 | 47.68 |
| Total indirect effect | 0.238 | 0.02 | 0.198 | 0.279 | 52.54 |
| EHL → SE → HR | 0.104 | 0.014 | 0.077 | 0.134 | 22.96 |
| EHL → SS → HR | 0.068 | 0.012 | 0.046 | 0.092 | 15.01 |
| EHL → SE → SS → HR | 0.065 | 0.009 | 0.049 | 0.084 | 14.35 |
Standardized effects are presented; 5,000 bootstrap samples; EHL, eHealth literacy; SE, self-efficacy; SS, social support; HR, health responsibility. The standardized coefficients in Table 4 and Figure 2 are based on ordinary least squares (OLS) regression using the full original sample, whereas the results in this table are derived from 5,000 bootstrap resamples. The minor difference of 0.001 between the two estimates reflects differences in computational logic, random sampling variation, and decimal rounding rules across software, which is a normal occurrence in mediation analysis.
Each contribution rate was calculated by dividing the individual pathway effect by the total effect. Because each value was rounded independently, the sum of contribution rates may not equal exactly 100%, which is a common statistical occurrence.