Three diagnostic approaches to Asperger syndrome: implications for research.
Different Asperger checklists pick different kids, so always name the exact tool you used.
01Research in Context
What this study did
Klin et al. (2005) ran three popular Asperger syndrome checklists on the same group. They wanted to see if the tools picked the same people.
The team looked at IQ scores, other diagnoses, and family patterns that each system found. They checked if switching rules changed the final head-count.
What they found
The three systems rarely agreed. Each rule set flagged different kids as having Asperger syndrome.
The tools also gave different IQ ranges and comorbidity rates. Family history looked different under each set of rules.
How this fits with other research
Evers et al. (2020) later saw the same mess with parent interviews. Only 16% of autistic kids passed both the 3di and DISCO-11 cut-offs, echoing the 2005 warning.
Richardson et al. (2008) compared the M-CHAT and SCQ in toddlers and again found that choice of tool changes who gets flagged. The pattern holds across ages.
Samyn et al. (2015) showed questionnaires and neuropsych tasks measure different things. The theme is clear: swapping forms is not a neutral move.
Why it matters
If you write a report, say which checklist you used. A child who meets DSM-IV-TR Asperger criteria may vanish under DSM-5 rules. The same is true for parent interviews and toddler screens. Pick one system and stick with it across re-evaluations. Your replication may fail if you quietly switch.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Open your last autism report and add one line that states the exact diagnostic system or interview form you used.
02At a glance
03Original abstract
OBJECTIVE: To examine the implications for research of the use of three alternative definitions for Asperger syndrome (AS). Differences across the three nosologic systems were examined in terms of diagnostic assignment, IQ profiles, comorbid symptoms, and familial aggregation of social and other psychiatric symptoms. METHOD: Standard data on diagnosis, intellectual functioning, comorbidity patterns, and family history were obtained on 65 individuals screened for a very high probability of having autism without mental retardation (or higher functioning autism, HFA) or AS. Diagnoses of AS were established based on three different approaches: DSM-IV, presence/absence of communicative phrase speech by 3 years, and a system designed to highlight prototypical features of AS. RESULTS: Agreement between the three diagnostic systems was poor. AS could be differentiated from HFA (but not from PDD-NOS) on the basis of IQ profiles in two of the three systems. Differences in patterns of comorbid symptomatology were obtained in two of the three systems, although differences were primarily driven by the PDD-NOS category. Only one of the approaches yielded differences relative to aggregation of the "broader phenotype" in family members. CONCLUSIONS: Diagnostic assignments of AS based on three commonly used approaches have low agreement and lead to different results in comparisons of IQ profiles, patterns of comorbidity, and familial aggregation of psychiatric symptoms across the approach-specific resultant groups of HFA, AS, and PDD-NOS.
Journal of autism and developmental disorders, 2005 · doi:10.1007/s10803-004-2001-y