Abstract
When presenting an automated aid to a user, there is variation in how much a human or the automation is involved in making and acting on the decision. Level of automation describes the balance between human and computer in making decisions and selecting actions. In this study, we presented subjects with a speeded length-judgment task at two different levels of automation, manipulating the order of the automation level between subjects. We found a significant effect that subjects who were presented first with the condition where the automated aid allowed them to directly select a response (i.e., lower automation level) responded faster in later trials when then presented with the condition where the automated aid required them to agree or disagree with the aid’s response (i.e., higher automation level). This effect was not found when the higher level of automation was presented first.
Get full access to this article
View all access options for this article.