Step 4 of the replication code: cleans the raw endline survey sections and saves the intermediate endline datasets (demographics, household, household composition, knowledge, practices, observations, and the decision-making, grit and Rosenberg scales).