An exploratory analysis of the use of cognitive, adaptive and behavioural indices for cluster analysis of ASD subgroups.
Add a brief ASD behavior rating scale to every assessment—clusters pop out and guide treatment better than IQ alone.
01Research in Context
What this study did
Richardson et al. (2008) ran a cluster analysis on 53 children with autism.
They mixed IQ scores, daily-living skills, and parent behavior ratings.
Math grouped the kids into clear sub-types, not just one big spectrum.
What they found
The numbers formed tight knots instead of a smooth line.
Each knot had its own mix of smarts, skills, and everyday actions.
Adding a behavior scale sharpened the picture more than IQ alone.
How this fits with other research
Rivard et al. (2026) tripled the sample and got the same three preschool knots, so the pattern holds.
Bitsika et al. (2018) later saw only two knots—high versus low severity—and said extra behavior tags were not needed.
The gap is age: V et al. looked at wider ages, while Vicki focused on school kids where severity swamps detail.
Gardner et al. (2009) used parent interview data and still found four knots, proving the idea works across tools.
Why it matters
If you test only IQ and adaptive scores you miss daily quirks that drive meltdowns or motivate learning.
Drop a quick parent behavior scale into your battery, then eyeball where the child lands in the cluster map.
That snapshot tells you which skills to target first and which triggers to plan for.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Hand the parent a 10-item behavior scale, score it with the IQ and VABS, and see which cluster the child matches.
02At a glance
03Original abstract
BACKGROUND: Autism spectrum disorder (ASD) includes symptoms that vary in severity and frequency between children. Consequently, multiple psychometric assessment procedures are recommended to detect an ASD, including scales which asses the presence and nature of ASD-related behaviour. However, to date, few studies have examined the relative and specific contribution which such behaviourally oriented scales can make to the diagnosis of individual children with ASD. METHOD: Cluster analysis was used to explore the common characteristics of a group of 53 preschool and elementary school children with an ASD, based upon scores on tests of cognitive ability, adaptive behaviour and behavioural checklists designed to measure the presence of typical ASD. RESULTS: Data confirmed the expected variability in intelligence test scores. In addition, measures of adaptive behaviour and data from a behaviourally based rating scale suggested that children with Asperger's Disorder and autism might be reclassified into subgroups according to the presence of particular ASD-specific behaviours. CONCLUSIONS: As well as confirming that children with ASD vary in their adaptive behaviour and cognitive levels, these data emphasise the limited contribution such assessment procedures make to an understanding of the child's day-to-day behaviour and functioning, thus arguing for the inclusion of behaviourally based rating scales to develop ideographic intervention plans.
Journal of intellectual disability research : JIDR, 2008 · doi:10.1111/j.1365-2788.2008.01123.x