This study used cross-measure factor analysis combining items from the Dimensional Assessment of Repetitive Behaviors (DARB) and Sensory Experiences Questionnaire (SEQ-3) to test whether sensory features and restricted/repetitive behaviors (RRBs) are best represented as a single shared domain — as grouped in DSM-5 — or as distinct latent dimensions.
The DSM-5 four-factor RRB-aligned model showed the **poorest** fit among first-order models. A 10-factor empirically derived model fit best. For bifactor models, a two-factor exploratory solution — Factor 1 (insistence on sameness, restricted/unusual interests, repetitive language, OCD behavior, self-injury, sensory hypersensitivity) and Factor 2 (repetitive motor behavior, sensory seeking, sensory hyposensitivity) — had the highest internal consistency and explained the most variance in clinical variables.
- Sample size and composition are not reported in the abstract, limiting generalizability assessment. - All measures rely on informant/caregiver report, which may introduce rater bias. - Cross-sectional design precludes evaluation of developmental changes in RRB/sensory structure.
Clinicians should not treat sensory features as a simple subdomain of RRBs; the data support at least two distinct dimensions that may warrant separate clinical attention and measurement. Future DSM revisions may need to restructure the RRB criteria to reflect this empirical split.
Explore related topics
