Abstract
Previous findings indicate that operator’s reliance towards a static automated aid increases with the degree of automation (DOA) especially when decision-making is affected. In this data reexamination of a previously conducted study, operators’ automation verification was investigated comparing a static automation supporting decision selection with an automation which in most trials only narrowed down possible diagnoses. Thus, in the majority of trials information sampling was essential for task completion in the latter condition. However, in a few trials the automation provided a diagnosis, too – giving participants the rare opportunity to fully rely on the automation. The question was investigated how participants behave in the exceptional occasions in which reliance is possible compared to participants who always have the opportunity to rely. Results show that when reliance was possible as an exception, participants verified their aid significantly less compared to the group who could rely throughout all trials. Implications for approaches of flexible automation are discussed.