Proceedings of the Annual Meeting of the Cognitive Science Society

Salient features of a task play an important role in how people create task representations which then influence strategy selection for accomplishing the task. We examined two theories, Represent-Construct-Choose-Learn (RCCL) and Rational Metareasoning (RM), both of which incorporate task features into their models of strategy selection. RCCL theory posits that when a strategy's success rate is low, it indicates that the task representation is not useful and those represented features are irrelevant in this case so people tend to drop these features from the task representation. Conversely, RM theory posits that strategy selection is based on consideration of all available features, with no discrete changes in the features incorporated into the task representation. A study was conducted to examine how participants changed their strategy choices based on the success rate of using a specific task feature. The results showed that neither theory aligned closely with empirical data.