Two-sided double-rank test for Fisher's sharp null hypothesis in a cluster-level proportional treatment effect model
Useful links