Modern Psychological Studies
University of Tennessee at Chattanooga
Place of Publication
Complex environments with automated systems, such as aircraft cockpits and nuclear control rooms, require critical decisions to be made about human intervention. Human monitors operating in these roles must interact with copious amounts of information. Decision support tools within integrated displays, especially alarms, aid people in monitoring these systems by capturing their attention to focus on possibly dangerous conditions. Once signaled, monitors choose whether they wish to acknowledge the alarm and search for more process status, or ignore it. This study investigates the impact of likelihood alarm technology versus traditional binary alarms on decision making accuracy and response bias in this acknowledgement phase using a two-stage Signal Detection Model. Participants performed two low-fidelity, twenty-min flight missions consisting of dual primary tasks, compensatory tracking and fuel management, and a secondary engine-monitoring task. Probability of engine malfunctions (10%, 90%) and type of alarm system (Binary vs. Likelihood) were manipulated for each participant. It was hypothesized that the probability of engine malfunctions (P), and likelihood alarm technology (LAT) would interact with decision making accuracy. Additionally, a main effect of P on decision making bias was expected. Results showed that LAT significantly increased accuracy, especially under low P conditions, but had no effect on response bias. The results of this study support prior literature's findings on the superiority of LAT over binary alarms in complex tasks characterized by high workload, translating to better monitoring performance for many practical applications.
BF1 .M63 v. 14 no. 1 2008
Clark, Rylan M. and Bustamente, Ernesto A.
"Enhancing decision making by implementing likelihood alarm technology in integrated displays,"
Modern Psychological Studies: Vol. 14
, Article 5.
Available at: http://scholar.utc.edu/mps/vol14/iss1/5