sketerpot comments on Error detection bias in research