Table 3.
Action theory test, conceptual theory test and significance of the mediated effect on physical activity (step count) – Baseline to 3-months (April - June, 2011) and baseline to 6-months (April - December, 2011) Australia
| Hypothesized mediators | Time |
Action theory test^ |
|
Conceptual theory test |
|
Direct effect |
|
Indirect effect |
95% CI |
Proportion (%) |
|---|---|---|---|---|---|---|---|---|---|---|
| A (SE) | p-value | B (SE) | p-value | C’ (SE) | p-value | AB (SE) | AB/(C’ + AB) | |||
|
Self Efficacy |
1 |
0.04 (0.08) |
0.62 |
1012 (635) |
0.11 |
4368 (621) |
<0.001 |
41 (107) |
−107 to 357 |
1% |
| 2 |
0.18 (0.11) |
0.10 |
1081 (587) |
0.68 |
3168 (725) |
<0.001 |
190 (168) |
−9 to 717 |
6% |
|
|
Enjoyment |
1 |
−0.16 (0.11) |
0.14 |
860 (487) |
0.08 |
4233 (625) |
<0.001 |
−140 (138) |
−697 to 19 |
3% |
| 2 |
−0.23 (0.11) |
0.05 |
1002 (527) |
0.06 |
3775 (688) |
<0.001 |
−226 (168) |
−732 to 7 |
6% |
|
|
Social Support (Friends) |
1 |
0.01 (0.12) |
0.91 |
407 (446) |
0.36 |
4034 (605) |
<0.001 |
−90 (70) |
−89 to 247 |
2% |
| 2 |
0.21 (0.14) |
0.14 |
497 (426) |
0.25 |
3423 (703) |
<0.001 |
106 (125) |
−34 to 573 |
3% |
|
|
Social Support (Family) |
1 |
0.03 (0.11) |
0.82 |
612 (450) |
0.18 |
3911 (605) |
<0.001 |
16 (86) |
−112 to 266 |
<1% |
| 2 |
0.06 (0.12) |
0.64 |
516 (515) |
0.32 |
3402 (694) |
<0.001 |
29 (102) |
−100 to 411 |
<1% |
|
|
Social Support (Teacher) |
1 |
−0.12 (0.15) |
0.43 |
−257 (337) |
0.45 |
3937 (617) |
<0.001 |
32 (78) |
−49 to 357 |
<1% |
| 2 |
0.54 (0.17) |
<0.001 |
828 (369) |
0.03 |
3037 (714) |
<0.001 |
445 (242) |
77 to 1068 |
13% |
|
|
School Environment |
1 |
0.28 (0.07) |
<0.001 |
−605 (733) |
0.41 |
4037 (680) |
<0.001 |
−172 (187) |
−574 to 173 |
4% |
| 2 |
0.58 (0.09) |
<0.001 |
742 (723) |
0.31 |
2933 (836) |
<0.001 |
434 (459) |
−415 to 1507 |
13% |
|
| Multi-mediation (all) | 2 | 3009 (894) | <0.001 | 390 (570) | −658 to 1714 | 11% |
1= baseline to 3-month.
2= baseline to 6-month.
Note. Control and intervention groups were coded ‘0’ and ‘1’ respectively; A = estimate of unstandardized regression coefficient of treatment condition predicting change in hypothesized mediators; B = estimate of unstandardized regression coefficient of change in hypothesized mediators predicting change in physical activity behavior, AB = product of coefficients estimate, C’ = effect of treatment condition on physical activity behavior controlling for mediator effect, SE = standard error, 95% CI = asymmetric bias-corrected bootstrap 95% confidence interval.