Validation of the autism spectrum screening questionnaire in a total population sample.

Posserud et al. (2009) tested the Autism Spectrum Screening Questionnaire (ASSQ) in every 7- to young learners in a Swedish town. They mailed the 27-item form to parents and teachers. They set a cut-off score of 17 or higher as a red flag for ASD.

Kids who screened positive got full clinical exams. The team then checked how well the red-flag score matched the exam results.

Using both parent and teacher forms, the cut-off caught 91 out of the kids who truly had ASD. It also correctly cleared 86 out of the kids who did not have ASD.

In plain numbers: 9 out of 10 autism cases were found, and 6 out of 10 flagged kids really were on the spectrum.

Lavi et al. (2023) later pushed the same idea to older kids. Their ASKY screener works from with nearly the same accuracy, showing the parent-plus-teacher method still holds.

So et al. (2013) got similar results using a 10-item scale pulled from the CBCL and Teacher Report Form. They also found high accuracy when both adults filled out the form.

De Kegel et al. (2016) saw the opposite with the full CBCL. Their study showed poor accuracy and lots of false alarms. The difference: Pety used a short, autism-tuned subscale, while Alexandra used the whole CBCL profile. Same tool family, different precision.

If you screen school-age kids for ASD, use a brief, autism-focused checklist and collect both parent and teacher data. A cut-off of 17 on the ASSQ gives you strong confidence in your referrals. When only one informant is available, know that accuracy drops—Eggleston et al. (2018) showed single forms miss many cases. Keep the two-form rule in your intake packet.