Towards the real-time measurement of mental workload

Darryl G. Humphrey, Arthur F. Kramer, Erik J. Sirevaag

Research output: Contribution to journalConference articlepeer-review


The primary goal of this study was to explore the utility of event-related brain potentials (ERP) as real-time measures of mental workload. To this end, subjects performed two different tasks both separately and together. One task required that subjects monitor a bank of constantly changing gauges and detect critical deviations. Difficulty was varied by changing the predictability of the gauges. The second task was mental arithmetic. Difficulty was varied by requiring subjects to perform operations on either two or three columns of numbers. Two conditions that could easily be distinguished on the basis of performance measures were selected for the real-time evaluation of ERPs. A bootstrapping approach was adopted in which one thousand samples of n trials (n = 1 to 75) were classified using several measures of P300 and Slow Wave amplitude. Classification accuracies of 90% were achieved with 15 trials. Results are discussed in terms of potential enhancements for real-time recording.

Original languageEnglish (US)
Pages (from-to)1426-1430
Number of pages5
JournalProceedings of the Human Factors Society
StatePublished - 1990
EventProceedings of the Human Factors Society 34th Annual Meeting - Orlando '90 - Orlando, FL, USA
Duration: Oct 8 1990Oct 12 1990

ASJC Scopus subject areas

  • General Engineering


Dive into the research topics of 'Towards the real-time measurement of mental workload'. Together they form a unique fingerprint.

Cite this