Table 2.
Hierarchical linear model to estimate the intervention effect( 47 ) of the European School Fruit Scheme in North Rhine-Westphalia, Germany, on the fruit and vegetable consumption of primary-school children aged 6–11 years
| Model 0 | Model 1 | Model 2 | ||||
|---|---|---|---|---|---|---|
| β | 95 % CI | β | 95 % CI | β | 95 % CI | |
| Fixed part | ||||||
| Constant | 1·558 | 1·41, 1·70 | 1·275 | 1·10, 1·46 | 0·982 | 1·17, 0·80 |
| Intervention (yes) | – | – | 0·763 | 0·58, 0·95 | 0·773 | 0·59, 0·96 |
| Gender (female) | – | – | – | – | 0·495 | 0·31, 0·68 |
| Age (grand mean centred) | – | – | – | – | −0·013 | −0·12, 0·10 |
| At school for lunch (yes) | – | – | – | – | 0·055 | −0·16, 0·27 |
| Random part | ||||||
| Time level (Level 1; variance within children)* | 1·499 | 1·276 | 1·276 | |||
| % | 77·3 | |||||
| Child level (Level 2; variance between children)† | 0·327 | 0·437 | 0·382 | |||
| % | 16·9 | |||||
| Class level (Level 3; variance between classes)‡ | 0·112 | 0·083 | 0·082 | |||
| % | 6·3 | |||||
| −2×log likelihood | 3446·572 | 3359·429 | 3330·694 | |||
*
n time 998.
†
n children 499.
‡
n classes 34.