EEG Correlates of Signal Rate, Time in Task and Individual Differences in Reaction Time During a Five-stage Sustained Attention Task

Abstract

The EEG of 20 subjects was monitored continuously while they performed a sustained attention task in which each subject performed five conditions in different order. All conditions involved regular presentation of digits (stimuli) at the rate of one per two seconds. Subjects were required to respond to particular digits (signals) and signal ratios varied between 10 and 50% over the five conditions. There was a brief rest between each condition, and total task time (including periods of rest) was approximately 112 min. The Results were: (i) Mean reaction time to wanted signals increased following the first condition (independent of signal ratio) and was also longest for the 50% signal ratio condition (independent of order of presentation), (ii) Errors (false positives and misses) increased as a function of signal ratio but not as a function of order of presentation. Subjects with faster mean reaction time committed more errors. However overall absolute error rate was low (3.2% under the most extreme condition), (iii) EEG abundance for the lower measured alpha frequencies increased as the task progressed, and mean dominant alpha frequency, decreased, (iv) EEG abundance for the higher measured alpha frequencies increased as a function of signal ratio i.e. the higher the ratio, the higher the abundance, (v) Subjects with higher EEG abundance and lower mean dominant alpha frequency were faster (mean RT) than subjects with lower abundance and higher mean dominant alpha frequency, (vi) EEG trends were different for fast and slow subjects; fast subjects gave EEG effects for both signal ratio and time in task, whereas slow subjects showed only time effects. The overall findings for time in task and the decrease in the quality of performance with decreased arousal are compatible with previous models of vigilance (e.g. Mackworth 1969). The unexpected results for signal ratio and individual differences (decreasing EEG arousal with task complexity and greater individual efficiency) are interpreted in terms of (i) loading on short term memory requirements by between-condition interference effects, (ii) inhibition of preparatory motor response to militate against response anticipation, and (iii) re-evoked orienting responses. Independent evidence suggests that both short term recall and motor inhibition call for lowered arousal if performance is to be successful.