Effects of external factors on smoking index (multivariate hierarchical linear regression model)
| Predictor | Cross-sectional effects on smoking index at baseline* | Changes in the impact of predictor on smoking index over time† | ||
| γ Coefficient (SE) | p Value | γ Coefficient (SE) | p Value | |
| Community support index | −0.062 (0.014) | 0.001 | −0.005 (0.022) | 0.796 |
| Anti-tobacco media messages | 0.020 (0.015) | 0.182 | 0.048 (0.021) | 0.021 |
| Enrolment | −0.042 (0.022) | 0.022 | 0.045 (0.021) | 0.029 |
| School socioeconomic context | −0.108 (0.024) | 0.001 | 0.133 (0.023) | 0.001 |
| School TUPE implementation index | 0.020 (0.019) | 0.297 | −0.043 (0.023) | 0.064 |
Smoking index is the standardised sum of four cigarette and one cigar smoking measures (α=0.68)
School socioeconomic context is the standardised sum of parent education, API scores and percentage of students eligible for federal school lunch subsidies (α=0.91)
↵* γ zero-one (γ01) in equation 4.
↵† γ one-one (γ11) in equation 4.