Minimum detectable effect size for main outcomes (accounting for sample
design and clustering)
Accounting for sample design and clustering, what is the minimum detectable effect size for main outcomes. Specify the unit, standard deviation, and percentage.
Assuming power of 80% and alpha as 5%, the minimum detectable effect size (MDE) that we will be able to detect for each treatment is 0.11 s.d. Pooling both treatment arms together, the MDE that we will be able to detect is 0.089 s.d.
A maximum of 1,000 firms will be approached with in-person firm visits and a maximum of 1,000 firms will be approached with the SMS+ approach. The MDE of in-person firm visits compared with SMS+ approach is 0.13 s.d.