Validation Is Necessary but Insufficient

Risk Adjustment Methods in Quality Measurement—Reply

“…They make a valid observation that the article presumes basic technical competence in model building (eg, more parsimonious models are preferred, calibration should be assessed) and that no model should receive an A -level grade in the absence of technical competence. Nonetheless, it is important to appreciate a larger issue: technical competence is necessary but not sufficient for producing a useful model …”