Proceedings of the Annual Meeting of the Cognitive Science Society

Many theories of reasoning and many experiments presupposethat human ability is stable over time, and so people usuallydraw the same conclusion from the same premises. The as-sumption has hitherto had little or no empirical investigation.We therefore analyzed a study in which 20 participants drewtheir own conclusions to the 64 sorts of syllogisms on twooccasions separated by roughly a week. We report the na-ture of the changes in the participants’ conclusions includingtheir spontaneous improvement in logical accuracy, and use amodel-based program, mReasoner, to explain the results.