Carl Westin

6 Introduction native solutions for solving the conflict, including a “best-ranked” resolution based on a cost-value comparison. 34 Although small-scale testing with experienced controllers indicated an overall favorable reception of CORA, there were three notable issues. First, signs of hesita- tion in accepting conflict resolutions suggest another issue, one of trust. Controllers frequently investigated the quality of conflict resolutions and occasionally expressed doubt as to whether a conflict actually would be solved with the suggested resolu- tion. Second, controllers spent much time in searching through the list of alternative conflict resolutions in an attempt to find a solution that they preferred. 34 Third, the choice of resolution strategies was found to differ between Area Control Centers (ACCs) and the nationality of controllers (e.g., the preference for lateral resolutions by Lisbon controllers in contrast to vertical resolutions by Malm¨o controllers), sug- gesting that the algorithm has to be context sensitive. While initiatives such as CORA have modeled and tailored decision aiding automation after controller conflict resolution strategies, they have not explicitly linked automation-generated solutions to the individual preferences of the con- troller. As such, they have not been able to ensure complete harmony between controller and automation decision-making strategies. In contrast, controllers are generally assumed to be homogeneous in how they prefer to solve conflicts. Con- sequently, automated CD&R decision aids have typically been designed to fit the group rather than the individual. On the basis of individual differences in personali- ties 36 and cognitive styles 37 that influence how problems are approached and solved, however, it can be expected that CD&R automation sensitive to individual differ- ences in solution preferences would be beneficial to automation acceptance. 1-4 Research goal It is reasonable to hypothesize that controllers would be more prone to accept auto- mated advice if the automation appears to reason and solve conflicts in a way that is similar to the controllers themselves. This notion can be captured in the concept of strategic conformance, defined here as the degree to which automation’s solution and apparent underlying operations match those of the human. While “conformance” addresses the solution match for a problem between hu- man and automation, “strategic” refers to the apparent underlying strategies for reaching that solution. In CD&R, the “solution” can be considered to be the mea- sures taken to solve a conflict (e.g., vectoring aircraft A behind aircraft B). The “apparent underlying operation” is the reasoning and rationale (of the automation) that seem (for the controller) to underlie a given solution (e.g., we vector aircraft A behind B because aircraft A is slower than aircraft B). Since the controller would