Abstract or Table of Contents
We introduce a model of skill acquisition that strikes a balance between top-down and bottom-up control using a knowledge representation in which pre- and post-conditions are attached to actions. This model captures improved performance, defined as shorter solution times and lower error rates during the task. The bottom-up control element also allows the model to show increased flexibility to solve similar problems and robustness against unexpected events. In two experiments using a complex aviation task, we contrasted instructions that explicitly stated preand postconditions with conventional instructions that did not. The augmented instructions led to better and more robust performance than standard instructions, especially on problems that required transfer. The results of Experiment 1 were fit by our model, which was then successfully used to predict the results of Experiment 2.