Skip to main content
. 2020 Oct 22;20(12):1234–1240. doi: 10.1111/ggi.14067

Table 3.

Analysis of the total, direct and indirect effects in the sample

Bootstrapping (5000)
Product of coefficients Percentile 95% CI Bias corrected percentile 95% CI
Point estimate SE Z Lower Upper Lower Upper
Combined model (n = 990)
Total 0.433** 0.056 7.675 0.322 0.543 0.361 0.601
Direct 0.227** 0.051 4.429 0.127 0.328 0.170 0.390
Indirect 0.206** 0.032 6.462 0.143 0.268 0.148 0.261
X → M1 → Y 0.122** 0.024 5.026 0.074 0.169 0.075 0.149
X → M2 → Y 0.060** 0.018 3.388 0.025 0.094 0.037 0.125
X → M1 → M2 → Y 0.024** 0.007 3.362 0.010 0.038 0.002 0.029
Men (n = 272)
Total 0.652** 0.118 5.520 0.421 0.884 0.414 0.928
Direct 0.382** 0.095 4.027 0.196 0.567 0.172 0.629
Indirect 0.271** 0.068 3.962 0.137 0.405 0.157 0.411
X → M1 → Y 0.156** 0.054 2.877 0.050 0.262 0.080 0.260
X → M2 → Y 0.087* 0.034 2.516 0.019 0.154 0.024 0.191
X → M1 → M2 → Y 0.029 0.016 1.732 −0.004 0.061 −0.006 0.055
Women (n = 718)
Total 0.358** 0.065 5.548 0.233 0.487 0.284 0.566
Direct 0.178** 0.060 2.965 0.060 0.295 0.118 0.374
Indirect 0.180** 0.035 5.137 0.113 0.252 0.119 0.242
X → M1 → Y 0.110** 0.022 3.257 0.028 0.113 0.057 0.141
X → M2 → Y 0.048** 0.025 4.104 0.053 0.151 0.022 0.126
X → M1 → M2 → Y 0.022* 0.004 2.158 0.001 0.018 0.006 0.029

*P < 0.05, **P < 0.01. All the values are unstandardized. X1 = pain, M1 = perceived social support, M2 = self‐rated health, Y = depression.

CI, confidence interval.