Searching for the Presence and Absence of Features in Sustained Attention: The Effect of Discriminability and Display Size on Performance and Workload

The present study examined the effects of signal discriminability and display size when searching for the presence and absence of features on vigilance performance and mental workload. Specifically, two conflicting theories were investigated. Treisman and Souther (1985) contend that the presence of a feature is detected quickly and effortlessly while the absence of a feature requires slower, serial processing. Nagy and Sanchez (1990, 1992), on the other hand, argue that discriminability determines the type of processing employed such that easily discriminable signals pop out of the display while less discriminable signals require focused attention to detect. To test these ideas, 80 participants were required to monitor a computer display comprised of either 2 or 5 circles during a 40-min vigil. Half of the participants were instructed to detect the presence of a line in one of the circles, and the other half were instructed to respond when the line was absent from one of the circles. In addition, half of the lines used were easily discriminable while the other half were less discriminable. Participants were also asked to provide estimates of mental workload for the task. The results indicated that observers performed more poorly over time. Feature type, discriminability, and display size were also found to be important determinants of performance and workload. Observers searching for the absence of a line performed worse and rated this condition as more difficult than those searching for the presence of a line. Searching through larger displays was also rated as more difficult and resulted in worse performance. In addition, performance in the difficult discrimination condition was worse, and individuals judged this condition to be harder than the easy discriminability condition. These results are discussed with respect to the aforementioned theories and the practical implications of these variables to the vigilance situation.