A Comprehensive Computational Model of Sustained Attention

Daniel Gartenberg

Advisor: Tyler H. Shaw, PhD, Department of Psychology

Committee Members: Patrick McKnight, Greg Trafton, Glenn Gunzelmann

David J. King Hall, CENTEC
March 02, 2016, 01:00 PM to 10:00 AM

Abstract:

The vigilance decrement is the decline in performance over time that characterizes tasks requiring sustained attention.  Resource Theory proposes that the vigilance decrement is due to information processing assets that become depleted with use.  Resource theorists must thus identify these assets and the process of how resources are depleted and replenished.  The Microlapse Theory of Fatigue (MTF) identifies the resource that is depleted when performing a sustained attention task as the central executive attentional network.  The depletion of the central executive network resource results in microlapses or brief gaps in attention that prevent the perception and processing of information.  The MTF can explain various effects in the sustained attention literature regarding how resources are depleted.  However, the MTF alone cannot explain the event rate effect or the motivation effect because it does not include replenishment mechanisms that can occur during a sustained attention task.  To better understand the process of replenishment, participants were assigned to varying event rate and external motivation conditions in a novel paradigm that could measure the perceptual processing of a trial over time.  These stages of processing included the when participants looked at the first stimulus, looked at the second stimulus, and responded.  In Experiment 1, it was found that the vigilance decrement was more severe for faster event rates, consistent with Resource Theory and counter to the MTF.  In Experiment 2, the event rate effect was replicated, but unexpectedly, external motivation did not impact the vigilance decrement.  In both experiments it was found that for the stages of processing that involved looking at the stimuli, more slowing was found as event rate increased.  Additionally, more slowing was detected earlier in the processing of a trial than later.  These results supported the process of microlapses inducing the vigilance decrement due to not having enough time to perceive, encode, and respond to stimuli, as described by the MTF.  It was interpreted that the interaction between time-on-task and event rate was due to opportunistic breaks that occurred more frequently in slower event rate conditions.  The finding that more slowing occurred earlier in processing was interpreted as evidence for internal rewards related to learning impacting the speed of processing a trial.  To explain these findings, I propose the Microlapse Theory of Fatigue with Replenishment (MTFR) a process model similar to MTF, but that includes additional replenishment mechanisms related to opportunistic rest periods and internal rewards.  The Microlapse Theory of Fatigue with Replenishment (MTFR) closely correlates to the empirical data and is an important step forward in the effort to build a comprehensive model of sustained attention.