New tool could quickly and accurately diagnose kids with autism

Embargoed until: Publicly released: 2026-07-23 01:00

International

CC:0

US researchers trialling computerised adaptive tests - where the previous answer alters the following one in a test - for detecting autism spectrum disorder in kids say their tool could be a quick and accurate means of working out if a child or young person has autism. The team calibrated their tool using answers from over 10,000 kids and say it was able to recognise a diagnosis with an average of 13 questions, as opposed to the classic questioning, which requires upwards of 420. The team say adaptively administering small, statistically optimal subsets of test questions and procedures can yield comparable (or better) results while reducing the burden on the kid performing the test.

Attachments

Note: Not all attachments are visible to the general public. Research URLs will go live after the embargo ends.

Research JAMA, Web page The URL will go live after the embargo ends

Journal/
conference: JAMA Network Open

Research:Paper

Organisation/s: University of California, Los Angeles, USA

Funder: This study was supported by funding from the Semel Institute for Neuroscience and Human Behavior at UCLA to Dr Jacob and the Center for Health Statistics at the University of Chicago to Dr Gibbons.

Media Contact/s

Contact details are only visible to registered journalists.