Exploring the Performance Consequences of Target Prevalence and Ecological Display Designs When Using an Automated Aid

Human operators do not necessarily perform better when receiving assistance from an automated aid than without the automated aid. The current work explored the impact of integrating the automated aid with the task information in low prevalence conditions. Specifically, we compared displays where the automated aid was integrated with task information in general or with another visual decision support aid. Subjects performed a speeded judgment task with the assistance of an automated aid, varying in display type, difficulty, and prevalence. Results indicated that participants performed less efficiently with the automated relative to without and that there was no added benefit of the visual decision support in terms of response times. Both decision supports improved participant’s sensitivity over no support, which may be beneficial for weakening the performance consequences of the low prevalence effect. Unexpectedly, we found that participants’ performance with each display was strongly dependent on when they experienced each display. It is possible that participants might be using strategies that complement one display over another, depending on the condition they see first. Automated aids could be used in real-world contexts to alleviate the effects of low target prevalence; however, the effectiveness may depend on experience with other interfaces.