Answers are grounded in this paper and cite their sources.
Italian psychometric evidence on the RAADS-R has been limited. This case-control validation examined the reliability, structure, and measurement invariance of the Italian RAADS-R, its discrimination between autistic and non-autistic adults with candidate cut-offs and likelihood-ratio (LR) bands, subscale contributions, gender differences, and AQ convergence. Sample 1 comprised 376 autistic adults referred to a specialized adult-autism service and diagnosed by blind multidisciplinary assessment, compared with 376 self-declared non-autistic community adults; Sample 2 comprised 354 adults with self-reported previously established diagnoses and 354 comparison adults, matched on age within gender. Analyses covered reliability, CFA/bifactor/ESEM structure, invariance, ROC analysis with cross-sample threshold application, LR bands, a Group × Gender MANCOVA, and AQ convergence. Internal consistency was excellent (ω = 0.98); a strong general factor (ω h = 0.96–0.97) indicated that the DSM-5 subscales describe content rather than distinct dimensions. In Sample 1, AUC was 0.91 and the candidate cut-off of 92 gave sensitivity 0.89 and specificity 0.81; Sample 2 yielded 104 (sensitivity 0.94, specificity 0.77). Across samples, 92 favored sensitivity and 104 specificity. Four LR bands graded the score (≤ 70, 71–91, 92–136, ≥ 137). Autistic women reported higher sensory scores than men, with item-level invariance supported. RAADS-R and AQ totals correlated strongly ( r = .86). The Italian RAADS-R is a reliable screening aid within multidisciplinary assessment. Because accuracy was estimated against community adults, the cut-offs and LR bands are preliminary, require validation in consecutive clinical samples, and should not be used to rule out autism when clinical suspicion exists.
No takes yet. Share an insight, caveat, or question.
Tortora et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: