Journal of Autism and Developmental Disorders vol:41 issue:2 pages:196-212
To test the spectrum hypothesis—postulating that clinical and non-clinical samples are primarily differentiated by mean-level differences—, this study evaluates differences in parent-rated temperament, personality and maladjustment among a low-symptom (N = 81), a highsymptom (N = 94) ASD-group, and a comparison group (N = 500). These classic spectrum hypothesis tests are extended by adding tests for similarity in variances, reliabilities and patterns of covariation between relevant variables. Children with ASD exhibit more extreme means, except for dominance. The low- and high-symptom ASDgroups are primarily differentiated by mean sociability and internal distress. Striking similarities in reliability and pattern of covariation of variables suggest that comparable processes link traits to maladaptation in low- and highsymptom children with ASD and in children with and without autism.