Psychophysiological Measures and Performance Effects in an Adaptive Sustained Attention Task

This experiment examined the effect of a biocybernetic system that had been successfully used to control an adaptive automation compensatory tracking task (Freeman, Mikulka, Prinzel, & Scerbo, 1999; Pope, Bogart, & Bartolome, 1995). The present study applied this adaptive automation system to a vigilance task. Recordings of EEG activity were used to derive an index of task engagement, beta/(alpha+ theta). A training session was used to establish a baseline level of this index. Index levels above baseline reflected an increase in arousal, while levels below baseline indicated decreased arousal. Three groups (production, reception, and control) performed the task which presented one target per minute coupled with non-target event rates of 6, 20, and 60 per minute. The production group used a negative feedback methodology in which EEG index values reflecting increased arousal produced a lower non-event rate. A yoked (reception) group was given the same pattern of event rate shills as the production group. The control group was simply required to perform the task with the event rate changing at random. Production group participants showed an attenuated decrement in accuracy scores over time. These initial results suggest that the biocybemetic system may provide a method of reducing the traditional vigilance decrement.