Table 2.
Partitioning the mediated paths: Total effects, direct effects, and indirect effects.
| Primary model: Depression symptoms predict smoking | ||||||
|---|---|---|---|---|---|---|
| 95% CIb | ||||||
| Effect | Estimate | S.E.a | Estimate/S.E. | p-value | Low | High |
| Depression symptoms level to peer smoking level to smoking level | ||||||
| Total effect | .26 | .08 | 3.13 | .002 | .10 | .42 |
| Total indirect effectc | .33 | .07 | 5.09 | <.0001 | .21 | .46 |
| Specific indirect effectd | .33 | .07 | 5.09 | <.0001 | .21 | .46 |
| Direct effect | -.07 | .08 | -.95 | .34 | -.22 | .08 |
| Depression symptoms level to peer smoking level to smoking trend | ||||||
| Total effect | .13 | .05 | 2.48 | .01 | .03 | .23 |
| Total indirect effect | -.03 | .05 | -.50 | .62 | -.13 | .08 |
| Specific indirect effect | .08 | .02 | 3.35 | .001 | .03 | .13 |
| Direct effect | .16 | .06 | 2.44 | .02 | .03 | .28 |
| Depression symptoms trend to peer smoking trend to smoking trend | ||||||
| Total effect | .29 | .22 | 1.29 | .20 | -.15 | .72 |
| Total indirect effect | .75 | .24 | 3.11 | .002 | .28 | 1.23 |
| Specific indirect effect | .75 | .24 | 3.11 | .002 | .28 | 1.23 |
| Direct Effect | -.47 | .26 | -1.76 | .08 | -.98 | .05 |
| Alternative model: Smoking predicts depression symptoms | ||||||
|---|---|---|---|---|---|---|
| 95% CI | ||||||
| Effect | Estimate | S.E. | Estimate/S.E. | p-value | Low | High |
| Smoking level to peer smoking level to depression symptoms level | ||||||
| Total effect | .09 | .03 | 3.08 | .002 | .03 | .14 |
| Total indirect | .15 | .04 | 4.08 | <.0001 | .08 | .23 |
| Specific indirect effect | .15 | .04 | 4.08 | <.0001 | .08 | .23 |
| Direct effect | -.07 | .05 | -1.37 | .17 | -.16 | .03 |
| Smoking level to peer smoking trend to depression symptoms trend | ||||||
| Total effect | -.04 | .02 | -1.73 | .08 | -.08 | .01 |
| Total indirect effect | -.10 | .04 | -2.97 | .003 | -.17 | -.04 |
| Specific indirect effect | -.06 | .02 | -2.76 | .01 | -.11 | -.02 |
| Direct effect | .07 | .04 | 1.71 | .09 | -.01 | .15 |
| Smoking trend to peer smoking trend to depression symptoms trend | ||||||
| Total effect | .04 | .03 | 1.17 | .24 | -.02 | .10 |
| Total indirect effect | .18 | .06 | 2.97 | .003 | .06 | .30 |
| Specific indirect effect | .18 | .06 | 2.97 | .003 | .06 | .30 |
| Direct effect | -.15 | .07 | -2.14 | .03 | -.28 | -.01 |
Note: Only significant indirect effects are reported
a
Delta Method Standard Errors
b
95% confidence intervals containing zero are not significant
c
total indirect effect represents the total of all indirect effects, including the specific indirect effect presented
d
the specific indirect effect is the significant indirect effect. Non-significant indirect effects are not presented in this table for parsimony