Domains of the Childhood Autism Rating Scale: relevance for diagnosis and treatment.

The team looked at 1,400 CARS records from kids with autism or other delays.

They ran a factor analysis to see which groups of items hang together.

Kids were and came from clinics and schools across the U.S.

Three clear factors popped out: Social Impairment, Non-compliance/Emotion, and Deviant Speech/Relating.

A Social Impairment score of 26 or higher correctly flagged 78 % of kids with autism.

That single factor also dropped a little after treatment, while the other two stayed flat.

Haring et al. (1988) already showed the full CARS can separate autistic teens from non-autistic ones.

Matson et al. (1994) now zoom in and say you only need the Social Impairment slice for a quick screen.

Narzisi et al. (2013) got 90 % accuracy with the CBCL toddler form, but L et al.’s CARS shortcut works across wider ages.

He et al. (2019) used the same factor trick on the RBS-R in Chinese kids, proving the method travels across scales and cultures.

You can save time by scoring just the Social Impairment items first.

If the total is 26 or above, move straight to full ADOS/ADI-R.

If it’s below, autism is unlikely and you can rule it out 78 % of the time without more testing.