Item-Level Self-Report Patterns Associated With Schizophrenia in a Transdiagnostic Cohort
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Background. The UCLA CNP LA5c cohort includes self-report scales drawn from personality, temperament, impulsivity, and perceptual-aberration instruments. We examined which items in this mixed battery distinguished participants with schizophrenia from the rest of the transdiagnostic cohort. Earlier analyses pointed to planning, caution, orderliness, and anomalous bodily self-experience. Methods. We analysed 19 preselected questionnaire items in N=271 participants from the UCLA CNP LA5c cohort, including n=39 with schizophrenia. We used ridge-penalised logistic regression to derive a cross-fitted multivariable item score. For each item, we reported its univariate association with diagnosis, correlation with the score, ridge coefficient after mutual adjustment, and held-out permutation contribution. Results. Participants with schizophrenia scored higher on attention to detail, caution, planning, organisation/rule-setting, and anomalous bodily self-experience items. Quick or careless action, dislike of detailed plans, and preference for spending over saving aligned in the opposite direction. The cross-fitted item score separated the groups modestly (AUC 0.705; log loss 0.392). Several deliberation items had smaller or reversed conditional weights after mutual adjustment, consistent with shared wording and content. Conclusions. The signal was distributed across items from several instruments. No single item accounted for the result.
Fields of Study & Subjects
Schizophrenia research and treatment
Psychological Testing and Assessment
Mental Health and Psychiatry
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 28 March 2026
- Language
- EN
- Document type
- Preprint