Modern Piano

Why We Cannot Rely on Ourselves for Epistemic Improvement

“…As noted by Robyn Dawes, David Faust, and Paul Meehl, there are now 'nearly 100 comparative studies in the social sciences' such that, '[i]n virtually every one of these studies, the actuarial [that is, statistical] method has equalled or surpassed the clinical method, sometimes slightly and sometimes substantially'. 32 This is, of course, excellent news for every clinician interested in making accurate diagnoses and prognoses. Indeed, in light of the prevalence of bias, the limits of self-correction-that is, of correcting for bias after the fact, in the manner discussed in the previous section-and the presence of highly reliable prediction models, one self-binding measure that clinicians could take is that of committing to using prediction models, whenever such models are available for the relevant clinical judgments.…”