Comments on statistical validity.