Abstract
We conducted an experiment to obtain objective measures of responses to binary alerts and subjective reports of the effects of alerts. We focused on the effects of the diagnostic value of the alert. Alerts helped improve performance, and participants responded more strongly and faster to alerts with higher diagnostic value. However, they did not report a greater reduction in workload with these alerts. Results show partial dissociation between objective and subjective measures of responses to alerts; while compliance and reliance measures were consistent with reports of system reliance and rating, effects of mean screen time were inconsistent with reports of temporal demand, effort invested and task difficulty, and participants’ performance using the alert systems did not correspond with their satisfaction with their own performance. This dissociation should be considered when evaluating the effects of automation.
Get full access to this article
View all access options for this article.
