Minimum detectable effect size for main outcomes (accounting for sample
design and clustering)
We estimate the number of clusters needed for a given minimum detectable effect size (MRES).
We assume the MRES to be 0.17 standard deviations, as previous studies have found this effect size (see Boeve-de Pauw et al., 2020; Innes et al. 2012; Roberts, 2012). We further assume power of 0.8, a Rho (ICC) of 0.01, R2 of 0.5, and a non-response rate of 50%. This yields a total of 62 clusters.