Abstract
The ability to maintain attention varies across and within individuals, affecting subsequent memory. How consistent is sustained attention to and its impacts on memory for information from different perceptual modalities? Can sustained attention performance be improved with real-time task manipulations? To ask these questions, we first developed an auditory-visual continuous performance task (avCPT). Online participants (N=110) completed two sessions of the avCPT in which they saw trial-unique scene images paired with trial-unique sounds every 1200 ms (500 trials). In each session, participants were instructed to attend to either images or sounds and press a button in response to frequent-category stimuli (indoor or outdoor scenes; natural or manmade sounds) and withhold responses to the infrequent category (10%). Performance (A’) on auditory and visual avCPT sessions was positively related across participants (r=.49, p<.001), suggesting that sustained attention is consistent across perceptual modalities. Furthermore, attentional state influenced subsequent recognition memory for visual stimuli. Infrequent-category images to which participants correctly withheld response were better remembered than those to which they incorrectly responded (t(105)=2.98, p=.004) but this effect was not observed for sounds. Given this evidence of reliability within individuals, we next asked whether performance can be improved via personalized feedback based on pupil size, an index of attentional state. In a modified version of the task above, participants categorized indoor and outdoor scenes whose clarity (% visual noise) varied as a function of real-time changes in pupil diameter. Whereas increasing perceptual difficulty in a low attention state (smaller pupil diameter) would improve A’ if poor performance results from mindlessness, decreasing difficulty in a low attention state would improve A’ if poor performance results from overload. While consistency in auditory and visual avCPT performance illustrate the stability of sustained attention performance across modalities, impacts of real-time task manipulation may provide evidence of malleability.