Minimum detectable effect size for main outcomes (accounting for sample
design and clustering)
To compare two groups, with a total number of 70 clusters with 5 randomization sites and 25 farmers per clusters we can detect a small to medium effect sizes between 0.20 SD and 0.33 SD with 80% power. Considering the baseline values as suggested by the pilot study on CS in Benin (Yegbemey et al., 2021), the anticipated effects sizes this implies MDEs between 4,366 and 7,719 XOF/ha for labor, 305 and 500 kg/ha for yield. Considering that the average daily pay in agriculture is about XOF 2,000, the saving in term of labor can be between 5 to 10 days. These MDEs are realistic given the pilot experiment. Considering that attrition rate will not exceed 10% the final sample we hope to have up to 30 farmers (about 30% more) involved per village, implying up to 3,150 farmers.