A factor analytic study of the Autism Behavior Checklist.
The ABC total score is fine for screening, but its original subscales are fiction — use the 2002 five-factor structure or stick with the total only.
01Research in Context
What this study did
The authors ran a factor analysis on the Autism Behavior Checklist. They wanted to see if the test’s five printed subscales really hang together.
They worked with children who have autism spectrum disorder. The math pulled out a new five-factor picture that explains 80 % of the score spread.
What they found
The old subscales do not match the real data pattern. A fresh five-factor set fits better, so the old labels can mislead you.
The total ABC score still works for screening, but the parts inside it need new names.
How this fits with other research
Gerhardt et al. (1991) already found a three-factor fit and warned the printed subscales were shaky. McCarron et al. (2002) now adds more detail with a five-factor fix.
Christopher et al. (1991) looks like a clash — they kept five factors on the Aberrant Behavior Checklist. The fight fades once you see they studied a different tool for kids with ID, not ASD.
Lan et al. (2025) later showed the ABC total still beats other screeners in a Chinese sample. The new factor map from 2002 helps you trust the total score even while you ignore the old parts.
Why it matters
Stop using the 1980s subscale names when you explain ABC results to parents or team mates. Instead, plot the five empiric factors or just report the total. This keeps your language in line with the data and cuts confusion at planning meetings.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Cross out the old subscale labels on your score sheet and write in the five new factors from M et al. (2002) before the next parent review.
02At a glance
03Original abstract
The factor structure of the Autism Behavior Checklist (ABC) (Krug, Arick, & Almond, 1980a, 1980b), a 57-item screening instrument for autism, was examined on a sample of 383 individuals with autism spectrum disorders (i.e., autistic disorder, Asperger syndrome, and other autism-like conditions) aged 5-22 years. A five-factor model accounted for 80% of the total variance in the checklist. Thirty-nine of the 57 items had factor loadings of 0.4 or more, with 13 items loading on Factor 1, 11 items on Factor 2, 6 items on Factor 3, 5 items on Factor 4, and 4 items on Factor 5. No support was found for classifying the 57 items into the five subscales proposed by Krug et al. (1980a, 1980b) or for the three-factor solution suggested by Wadden, Bryson, and Rodger (1991).
Journal of autism and developmental disorders, 2002 · doi:10.1023/a:1015519413133