- One-factor RFQ-8 model (with a residual correlation between items 3 and 4) showed adequate fit and acceptable reliability in SSD and controls.
- Partial metric invariance held, but item 7 showed scalar non-invariance; reverse-coded items problematic in SSD, suggesting revision or removal.
- Emotional dysregulation was the strongest predictor of hypomentalising across groups; attachment effects differed, secure attachment predicted better mentalising only in controls.
J Pers Assess. 2026 Sep 2:1-10. doi: 10.1080/00223891.2026.2722651. Online ahead of print.
ABSTRACT
There is a need to effectively measure mentalization in people with Schizophrenia Spectrum Disorders (SSD) due to its relevance for delusions and hallucinations. The RFQ-8 appears to be an easy-to-administer screening tool for mentalization, but its factor structure is debated. In this study, we examined the psychometric properties of the RFQ‑8 model in individuals with SSD (n = 100) and healthy controls (n = 380). Criterion validity was tested through associations with attachment style and emotional dysregulation. Multiple group confirmatory factor analysis and structural equation modeling were used. With a single residual correlation between items 3 and 4, a one-factor solution showed adequate fit. Reliability was acceptable (ωcases = .75; ωcontrols = .76). Partial metric invariance held, but item 7 showed scalar non-invariance, suggesting it should be revised or removed. Secure attachment predicted better mentalizing only in controls, while avoidant and ambivalent styles related to reduced mentalizing. Emotional dysregulation was the strongest predictor of hypomentalizing across groups. Overall, the findings provide support for the single‑factor RFQ‑8 but indicates some measurement issues, particularly related to reverse‑coded items in SSD populations.
PMID:42683917 | DOI:10.1080/00223891.2026.2722651
Share Evidence Blueprint
Save to Google Notes
Search Google Scholar
Save as PDF
⭐ My Revision List
