Which Measures From a Sustained Attention Task Best Predict ADHD Group Membership?

Keitaro Machida, Edwina Barry, Aisling Mulligan, Michael Gill, Ian H. Robertson, Frances C. Lewis, Benita Green, Simon P. Kelly, Mark A. Bellgrove, Katherine A. Johnson

Research output: Contribution to journalArticleResearchpeer-review

6 Citations (Scopus)


Difficulty with sustaining attention to a task is a hallmark of ADHD. It would be useful to know which measures of sustained attention best predict a diagnosis of ADHD. Participants were 129 children with a diagnosis of ADHD and 129 matched controls who completed the fixed Sustained Attention to Response Task (SART). The number of commission and omission errors, standard deviation of response time (SDRT), tau, fast and slow frequency variability, d-prime, and mu were able to successfully classify children with and without ADHD. The mean response time, criterion, and sigma were not able to classify participants. The best classifiers were d-prime (0.75 Area Under the Receiver Operated Characteristic), tau (.74), SDRT (0.74), omission errors (0.72), commission errors (0.71), and SFAUS (0.70). This list of the best classifier measures derived from the SART may prove useful for the planning of future studies.

Original languageEnglish
Pages (from-to)1471-1482
Number of pages12
JournalJournal of Attention Disorders
Issue number11
Publication statusPublished - Sept 2022


  • ADHD
  • response time
  • SART
  • sustained attention

Cite this