This component includes the original WASH Benefits Kenya study protocol (Arnold et al. 2013) and a short update that pre-specifies some small changes and additional details in the trial's statistical analysis plan. All decisions were made fully blinded to the trial's randomization assignment.
**Replication and blinding**
The primary analysis -- including data management -- is conducted fully blind in duplicate by two independent team members, Dr. Jade Benjamin-Chung and Andrew Mertens. We will unblind the trial's randomization assignment only after independent primary analyses achieve perfect replication.