Screening accuracy of Level 2 autism spectrum disorder rating scales. A review of selected instruments.

Norris et al. (2010) looked at every paper that tested Level 2 parent checklists for autism. They pulled the numbers on how often each tool caught kids who really had ASD and how often it flagged kids who did not. The team compared four common forms: SCQ, SRS, ASSQ, GARS/GARS-2, and ASDS.

The SCQ came out on top: it usually caught autism and rarely cried wolf. SRS and ASSQ looked useful but had less proof behind them. GARS, GARS-2, and ASDS missed too many children, so the authors cannot recommend them.

Barnard-Brak et al. (2016) later showed the SCQ is not age-proof; accuracy drops when you use the same cutoff for different ages. This seems to clash with Megan’s “good accuracy” label, but the difference is method: Megan pooled all ages, while Lucy split ages and saw the fade.

Hampton et al. (2015) ran a meta-analysis that updated Megan’s map. They still liked parent screeners with strong social-interaction items, agreeing SCQ is solid while adding that broad-band tools can work too.

Bao et al. (2017) took the same question to low- and middle-income countries. They found the same forms in use, but scores swung widely when cultural tweaks were made, backing Megan’s warning that the evidence base is thin.

When you pick a screener, reach for the SCQ first, but recheck the age range of the cut-score you use. Do not trust GARS or ASDS to rule out autism. If you serve culturally diverse families, collect local norms before you trust any translated form.