Task-aligned outcome learning in psychiatry: reducing endpoint dilution (opens in new tab)

Psychiatric research relies on well-defined outcomes for standardization, comparability, and replication, yet investigators often fix broad endpoints before knowing which symptom domains carry task-relevant signal. Even when psychometrically sound and clinically useful, composite measures can dilute predictive information and attenuate treatment effects when predictability or responsiveness concentrates in only a subset of symptoms—thus making studies appear negative despite meaningful change...