Factors Affecting Performance Using Modified Sustained Attention to Response Tasks

Justin Mensen

Advisor: William S Helton, PhD, Department of Psychology

Committee Members: Matthew Peterson, Tyler Shaw

Online Location,
April 27, 2022, 01:00 PM to 03:00 PM

Abstract:

Since 1997 the Sustained Attention to Response Task (SART) has been used to measure human performance by evaluating speed and accuracy of participants’ responses. To expand on the existing research a series of three studies was conducted. The first study examined the effects of real-time feedback on SART performance and determined that while real-time feedback can alter participants performance to be either faster or more accurate, there is still an inherent Speed-Accuracy Trade-Off (SATO) regardless of the presence of feedback.  Since the study findings supported the assertion that the SART measures the SATO and not sustained attention or vigilance, I propose that a more accurate description of the task would be the Speed-Accuracy Response Task.  For the second and third studies, a spatial SART-task, based on the previous work of Dr. Helton and associates, was modified into an agricultural scenario in which  participants were tasked with identifying and spraying weeds and withholding spraying responses when they identified soybeans.  The second study incorporated a motor task requiring the subjects to move a cursor and click to spray.  The third study removed the motor task and only required subjects to click as the cursor was automatically centered on the target. The results of these two studies show that incorporation of a motor task, while increasing response times slightly, greatly increases the ability to successfully identify and withhold spraying soybeans.  However, when the third study removed the motor task, soybeans were much more likely to be incorrectly sprayed than in study two.  Combining the findings of these three studies provides useful information regarding the real world implications of the SATO as effected by the presence or absence of a motor task and real-time feedback for systems used by operators to quickly make Go/No-Go decisions for a variety of real-world applications.