Language-model ratings of depression reflect the rater more than the patient

arXiv:2610.08501v1 Announce Type: cross Abstract: Depression has no diagnostic blood test. Language models promise tireless, consistent assessment, but can accurate raters disagree about individuals? We pre-registered 880 language-model raters, crossing 11 open models with prompting and scoring choices, and applied them to 189 interviews against the eight-item Patient Health Questionnaire. Model choice explained 30.0% of summed-symptom score variance, stable participant differences 10.5%. Two randomly drawn raters with area under the receiver operating characteristic curve (AUC) >= 0.70 disagreed on screening decisions for 40% of participants, on average. Average over-rating governed how many were flagged, yet equal-capacity raters chose differently for about one participant in five. A locked analysis of 86 new interviews reproduced the main pre-registered findings. Exploratory recalibration with 40 labelled participants raised accuracy from about 60% to 75% and halved disagreement, leaving one participant in five decided differently. Calibration repaired much of the rater dependence without securing agreement about individuals.

Related

• AI Rater Discrimination Depends on Scoring Protocol in Complex Clinical Decision-Making
• Depression Risk Assessment in Social Media via Large Language Models
• Depression Symptoms and Relational Patterns in 187k ChatGPT Histories

Source: arXiv cs.AI | 2026-10-07