ABSTRACT
This chapter presents empirical evidence for two aspects of the ACT-R performance theory. It looks at conflict resolution, which is dealt with in a substantially different way than it was in the ACT* theory. The chapter concerns the microscopic analysis: conflict resolution among the rules. Subjects appeared to be choosing among moves in a way that correlated with expected value and selecting moves in a way that is correlated with the models satisficing process. The authors have performed two experiments looking at subject's problem solving with these progressive maps. The first experiment used relatively simple maps. The second experiment introduced a number of additional complications to further test the theory. The chapter presents a theoretical analysis appropriate to the simpler maps and the results from the first experiment. The production rules used to model this task are derivatives of the production rules that were used for the first experiment, but there are some modifications.
