Minimum detectable effect size for main outcomes (accounting for sample
design and clustering)
Reaching a total of 1,800 companies, we have the capacity to detect an effect of 0.15 standard deviations with two treatment groups (two alternative messages). This means we could detect the treatment effect if it resulted in at least an increase of 74 applications on average per job posting (with an average of 408 applications), and 58 applications from women (with an average of 263 applications from women).