3 ms·

Protocol violations invalidate study data by definition. Cherry picking the data would just bias it worse.