Answers are grounded in this paper and cite their sources.
Retrospective neurodevelopmental diagnosis (RND) rarely reports validity. On a database of 5,432 deceased historical figures, 237 anchors independently hold both a third-party Murray-style five-level achievement grade and a four-level psychopathology-severity rating, both assigned blind to the tool's six-dimensional Bayesian P_ND. Continuous convergent validity is weak but significant (anchor mean 0.274 vs 0.241; Welch t=2.935, p=.0036, d~0.13), whereas high-threshold disorder classification (P_ND>=.60) is not enriched (3.0% vs 2.2%; Fisher OR=1.346, p=.374). Discriminant validity: P_ND correlates only weakly with independent pathology severity (Spearman rho=.203, p=.0017) and the high-threshold rate is 0% in the severe group (vs 3.1% with none; Fisher p=.51); achievement class does not monotonically predict high-threshold ND (zero Class 1). The tool supports a small real continuous creativity-neurotrait association, rejects 'great people generally have ND' and 'ND is severe mental illness', and quantifies the eminent-sample collider. Valid only within well-documented comparable subgroups; labels are most-plausible hypotheses, not diagnoses.
No takes yet. Share an insight, caveat, or question.
Xin Xu (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: