Abstract
Two experiments attempted to improve the quality of people's probability assessment through intensive learning. The first showed considerable learning almost all of which was accomplished after receipt of the first feedback. The second revealed the same pattern of learning and limited generalization. The implications of these results are discussed.