Table 6.
The effect of a 2-point change in the SDQ subscales on non-completion of upper secondary school, and the mediating effect of grades
| |
Model 1
a
|
Model 2
b
|
Model 3
c
|
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |
Total effect |
Proportion via mediation |
Total effect |
Proportion via mediation |
Total effect |
Proportion via mediation |
||||||||||||
| Estimate | 95% CI | Estimate | 95% CI | Estimate | 95% CI | Estimate | 95% CI | Estimate | 95% CI | Estimate | 95% CI | |||||||
|
Boys |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| SDQ externalising |
0.068 |
0.059 |
0.077 |
0.702 |
0.611 |
0.801 |
0.069 |
0.060 |
0.079 |
0.714 |
0.629 |
0.813 |
0.047 |
0.038 |
0.058 |
0.831 |
0.692 |
1.032 |
| SDQ internalising |
0.021 |
0.010 |
0.031 |
0.540 |
0.324 |
0.938* |
−0.00056 |
−0.010 |
0.010 |
0.428 |
−12.90 |
14.00 |
0.0031 |
−0.006 |
0.013 |
−0.740 |
−14.20 |
22.90* |
|
Girls |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| SDQ externalising |
0.067 |
0.057 |
0.077 |
0.716 |
0.628 |
0.833* |
0.063 |
0.053 |
0.073 |
0.732 |
0.624 |
0.860* |
0.044 |
0.033 |
0.055 |
0.755 |
0.610 |
0.976* |
| SDQ internalising | 0.031 | 0.023 | 0.039 | 0.622 | 0.488 | 0.823* | 0.012 | 0.005 | 0.020 | 0.412 | 0.185 | 0.952* | 0.012 | 0.005 | 0.020 | 0.310 | 0.108 | 0.687* |
aModel 1: Adjusted for ethnic background, county of residence, parents education, income and marital status.
bModel 2: SDQ externalising and SDQ internalising adjusted for each other, and for ethnic background, county of residence, parents education, income and marital status.
cModel 3: Like model 2, in addition adjusted for physical activity, smoke and alcohol habits.
*Direct effect statistically significant (p < 0.05).
The total effect estimate is the change in the probability of non-completion.