The Influence of Affective Valence and Task Uncertainty on Visual Threat Detection and Automation Use

Date of Award

Summer 2014

Document Type


Degree Name

Doctor of Philosophy (PhD)




Human Factors Psychology

Committee Director

P. Madhavan

Committee Member

J. Bliss

Committee Member

G. Watson


Objective: The purpose of this study was to examine the influence of affective valence on signal detection performance and automation use under different levels of uncertainty and with different levels of automation reliability. Method: In Experiment 1, participants completed an affect induction procedure designed to evoke positive, negative, or neutral affect. Participants then completed a visual Synthetic Aperture Radar (SAR) screening task where they were tasked with detecting the presence of foes among friends on the battlefield. The clarity of the SAR images was manipulated such that half of the images (100 per block) were relatively easy to distinguish (low uncertainty) and the other half were difficult to distinguish (high uncertainty). During this screening task, half of the participants received the assistance of an automated decision aid called Detector which was designed to be 100% accurate. Overall, 109 participants (37 males and 72 females, age range: 18-51) completed three affect manipulations (all of the same valence), 300 low uncertainty detections, and 300 high uncertainty detections. Participants were assessed for system trust and perceived reliability, automation compliance and reliance, self-confidence, and detection performance. Experiment 2 was almost identical to Experiment 1 except that Experiment 2 was shortened to two blocks and all participants received the assistance of an imperfect Detector (70% reliable vs. 90% reliable). Thus participants received a total of two affect manipulations, 200 high uncertainty images, and 200 low uncertainty images. One hundred thirty-one participants (45 males and 86 females, age range: 18-33) completed the study. Results: Data were analyzed using 3 (affect) x 2 (automation) x 2 (task uncertainty) mixed ANOVAs with follow-up tests as appropriate. In Experiment 1, low uncertainty led to higher sensitivity and reported confidence than high uncertainty. Affective valence interacted with automation presence to impact criterion setting and affect interacted with uncertainty and automation presence to influence confidence. In Experiment 2, low uncertainty led to higher sensitivity, confidence, and reliance than high uncertainty. Automation reliability influenced trust, perceived reliability, and dependence. Conclusion: Results reveal that complex visual search performance is impacted by automation reliability, task uncertainty, and to a lesser extent, operator affective valence.


In Copyright. URI: This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).





This document is currently not available here.