Carl Westin

7-2 Explaining the acceptance of conformal advisories 157 controllers’ solution objectives (i.e., the preferred classification) did vary between biased and unbiased conflicts. If controllers perceived repetitions differently, they are also likely to have fo- cused on different information cues when determining a solution. The dynamic simulation environment made scenario repetitions susceptible to changes beyond control as controllers freely could interact. Although scenarios were identical “on paper,” the interactions made during simulations made each repetition unique. This can also explain why controllers were not consistent down to the more detailed de- cision stages. Moreover, some learning effects can be expected since solution feed- back always was provided in that scenarios continued after the conflict had been solved. This is likely to have also influenced controllers’ evaluation, strengthening good solutions and weakening poor solutions. 7-2 Explaining the acceptance of conformal advisories As always when studying people, it is not how we as researchers have defined the world that matters; What matters is how the world is perceived by the people being studied. Although the exact same scenario and conflict were repeated, controllers may have perceived them as different. Taken together, results suggest that con- flict solving is predominantly intuitive, characterized by an instinctive, automatic, and reactionary decision-making process. This process is represented by several decision-making models that have been discussed throughout this thesis, such as recognition-primed decision-making 262, 263 in naturalistic environments, 264 System 1 decision-making, 41 and the peripheral route of the elaboration likelihood model (ELM). 265 They all capture a process whereby decisions are based on intuitive ex- pertise 21, 266 driven by an instinctive (less rational) reaction triggered by the recog- nition of familiar information cues and their relationship. In ATC CD&R, these patterns have typically been described in terms of heuristic decision-making pro- cesses, subjectively manifested as rule of thumbs such as “vectoring behind” and “vector away from conflict.” Based on observations and findings, Figure 7-2 illustrates a proposed model of the underlying decision-making process of a controller in relation to the confor- mance of a resolution advisory. The model builds on the lens model 267 (the percep- tion of cues in the upper part of the model) and recognition-primed decision model (the decision-making process in the lower part of the model). 263 The conflict situ- ation (top of the model) represents the objective ground truth of the conflict. The conflict is defined by a several mathematical and geometrical objective parameters . The model initially illustrates two different paths describing conflict identifica- tion, separately for a controller and an automated CD&R decision aid. In relation