Minimum detectable effect size for main outcomes (accounting for sample
design and clustering)
For school recruitment, the schools were contacted through the municipalities. Subsequently, with the database of schools available for the study, we performed randomization with the statistical program STATA to generate a treatment and control group of 2 schools in each experimental group. The number of schools needed for this study was calculated based on estimates of observations needed to observe statistical differences using the G*Power program. This sample size was calculated considering that repeated measures analysis of variance (ANOVA) with inter- and intra-subject factors will be performed, considering an effect size f 0.2, an alpha error of 0.05, a power of 0.95, two groups (intervention and control) and two measurement moments per group (pre and post intervention). Only two measurements were considered in the calculation, given that the study seeks to provide knowledge from the first post-intervention assessment, given the relevance of finding effects that are not only statistically significant, but also scientifically distinctive, and then follow the effect of the intervention over time. Considering that in most intervention studies there is loss of participants for various reasons (in educational context e.g. they change educational establishment), we considered recruiting 10% more than the calculated sample size (Gupta et al., 2016), reaching a final recruitment number of 94 children in fourth grade (9 to 10 years old).