People become more logical without feedback

Most theories of reasoning assume that human reasoning ability is stable over time, but few studies have investigated the matter. We analyzed a study in which 20 participants drew their own conclusions to the 64 sorts of syllogisms on two occasions separated by roughly a week.

In the latest CogSci proceedings, we report the nature of the changes in the participants’ conclusions including their spontaneous improvement in logical accuracy, and we show how a model-based computational cognitive theory, mReasoner, can explain the results.