Abstract
Presentation of an aid's reliability may have mixed results on utilization due to differences in the way the reliability is framed, as well as its effects on reliance versus compliance. Using an aided signal detection task, the current study compared effects of attribute framing (80% correct versus 20% incorrect) and goal framing (focusing on hits and correct rejections versus misses and false alarms) on sensitivity, response bias, compliance, and reliance. Sensitivity and response bias results indicated few differences between aided groups. Information framing did lead to differences in compliance, but not reliance compared to an unframed group. However, reliance differences occurred when comparing positive and negative framing group. These results indicate that the manner in which information about the reliability of a diagnostic aid is presented can have significant, albeit subtle, effects on automation utilization.
Get full access to this article
View all access options for this article.