Automated aids provide users additional information for making decisions. The way the aid presents the information requires the user to either make the same decision as unaided or to agree or disagree with the aid’s recommendation. In this study, we measured response times and accuracy without an aid and with an aid where either: 1) the subject makes the same decision as the unaided condition, or 2) the subject agrees or disagrees with the automated aid’s decision. Results show subjects were more accurate with direct selection decisions, more accurate aids, and easier tasks, with an interaction between decision type and aid accuracy. Subjects were faster with direct selection decisions and more accurate aids, with an interaction between decision type and aid accuracy. Using a cognitive model we found information accumulation rates and caution varied across conditions.
& Houpt, J.
(2019). Effects of Decision Type and Aid Accuracy on User Performance. 20th International Symposium on Aviation Psychology, 361-366.