Thinking, Fast and Slow · chapter 22 · id thinking-fast-and-slow-c22-02-simple-statistical-models-cons

“Simple statistical models consistently outperform expert clinical judgment in structured prediction tasks, as demonstrated by Meehl (1954) and confirmed by subsequent meta-analyses.”

holdsconfidence: high✓ ZHP-verified + red-teamed

Receipts

Grove, Zald, Lebow, Snitz & Nelson (2000), Psychological Assessment 12(1)DOI registry: valid

Across 136 studies, mechanical (statistical) prediction substantially outperformed clinical prediction. The advantage held across medicine, psychology, and education. Only a small minority of cases showed clinical judgment outperforming the formula.

Aegisdottir et al. (2006), The Counseling Psychologist 34(3)FLAGGED: DOI not in registry

Fifty-six years of accumulated research confirms clinical judgment is generally inferior to statistical prediction in structured, repeated prediction tasks.

This claim is a stable, citable object. If you can falsify a verdict, tell us — corrections are loud here.