Table 3.
Results for the model testing mean-level differences between conditions on water and SSB consumption after the intervention (n 451)
| Water consumption | SSB consumption | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| b | se | β | P | CI | b | se | β | P | CI | |
| Regression paths | ||||||||||
| Active control [1] v. SNI [0] – Behaviour T2 | –0·02 | 0·18 | –0·01 | 0·936 | −0·20, 0·19 | –0·03 | 0·06 | –0·04 | 0·633 | −0·18, 0·11 |
| Control [1] v. SNI [0] – Behaviour T2 | –0·17 | 0·31 | –0·09 | 0·589 | −0·43, 0·24 | 0·20 | 0·10 | 0·25 | 0·035 | 0·02, 0·48 |
| Active control [1] v. SNI [0] – Behaviour T3 | –0·31 | 0·25 | –0·16 | 0·217 | −0·40, 0·09 | 0·17 | 0·10 | 0·20 | 0·061 | −0·01, 0·40 |
| Control [1] v. SNI [0] – Behaviour T3 | –0·20 | 0·20 | –0·10 | 0·308 | −0·29, 0·09 | 0·19 | 0·13 | 0·21 | 0·131 | −0·06, 0·49 |
| Stability paths | ||||||||||
| Behaviour T1 – Behaviour T2 | 0·51 | 0·06 | 0·47 | < 0·001 | 0·36, 0·58 | 0·51 | 0·06 | 0·49 | < 0·001 | 0·32, 0·66 |
| Behaviour T2 – Behaviour T3 | 0·62 | 0·05 | 0·58 | < 0·001 | 0·51, 0·64 | 0·62 | 0·05 | 0·64 | < 0·001 | 0·46, 0·82 |
| Control variables | ||||||||||
| Sex* – Behaviour T2 | –0·00 | 0·16 | –0·00 | 0·977 | −0·08, 0·08 | –0·00 | 0·16 | –0·08 | 0·119 | −0·18, 0·02 |
| Sex* – Behaviour T3 | 0·02 | 0·24 | 0·01 | 0·923 | −0·11, 0·12 | 0·02 | 0·24 | –0·03 | 0·431 | −0·10, 0·04 |
SSB, sugar-sweetened beverage; b, unstandardised regression coefficient estimating the mean-level difference between conditions, adjusted for previous consumption; β, standardised regression coefficient; SNI, social network intervention; T1, Time 1; T2, Time 2; T3, Time 3.
*
0 = boy, 1 = girl; boldface indicates statistical significance (P < 0·05); numbers in parentheses represent the binary dummy-coded values; SNI is the reference category in the model.