Crystal Smit

Chapter 2 36 intentions independent-samples t -tests were conducted. Pearson’s chi square tests were performed to check whether there were differences in sex. Table 2.1 summarizes the means and standard deviations ( SD s) for all variables across the conditions. No differences ( p > .05) were found between the intervention and control condition, which indicated that the randomization was successful. Table 2.1 Randomization checks of the variables measured by control and intervention condition 1 Intervention ( n = 106) Control ( n = 104) Total ( N = 210) p -value 2 Age (y) 10.67 9 ± - .78 13 10.83 9 ± - .82 13 10.75 9 ± - .80 13 .16 Boys/girls (n/n) 51 / 55 50 / 54 101 / 109 Thirst 2.11 1.00 ± - .79 4.00 2.09 1.00 ± - .73 4.00 2.10 1.00 ± - .75 4.00 .80 Water consumption 2.67 0.00 ± - 1.40 5.00 2.37 0.00 ± - 1.39 5.00 2.52 0.00 ± - 1.40 5.00 .11 SSB consumption 1.28 .17 ± - .84 5.00 1.18 .17 ± - .60 2.83 1.23 .17 ± - .74 5.00 .30 Water drinking intentions 2.83 1.00 ± - .75 4.00 2.75 1.00 ± - .74 4.00 2.79 1.00 ± - .74 4.00 .49 Note. 1 Values are presented in means ± SDs, min. - max. 2 Reflects the differences in total means between the conditions by independent-samples t -tests or Pearson’s chi square test. Main Analysis Pearson’s correlation analyses with all the dependent variables in the overall sample showed that thirst correlated significantly at both baseline and post- intervention with water consumption ( r baseline = .27, p < .001; r post-intervention = .23, p = .001) and water drinking intentions ( r baseline = .15, p = .032; r post-intervention = .23, p = .001). However, thirst only correlated marginally significant with SSB consumption at post-intervention ( r baseline = .09, p = .199; r post-intervention = .13, p = .052). This was not the case for age and sex ( p > .05). To make sure that thirst did not confound the effects, we included thirst at both time points as covariates in the main analysis. To examine whether children in the intervention condition reported consuming more water, less SSBs, and stronger intentions to drink water post-intervention, compared to those in the control condition, a two-way repeated measures MANCOVA was performed. There was a significant multivariate interaction effect