dclayh comments on Error detection bias in research