The checks a monitoring team runs on a survey dataset before anyone estimates anything: duplicated and near-duplicated interviews, Benford's law on amounts, age heaping (Whipple and Myers), straightlining in rating batteries, interview length and timing, missingness, and an enumerator-by-enumerator profile that says who to call. Drop in a CSV; it never leaves your machine.
Data
or
The demo is simulated (seeded) with six planted problems, so every check has something to find. It is labelled as such throughout.
Overview
Load a dataset to begin.