摘要：The notorious bat-and-ball problem has long been used to demonstrate that people are easily biased by their intuitions. In this paper we test the robustness of biased responding by examining how it is affected by repeated problem presentation. Participants solved 50 standard and control versions of the bat-and-ball problem. To examine the nature of a potential learning effect we adopted a two-response paradigm in which participants have to give a first hunch and can afterwards take the time to deliberate and change their answer. Results showed that both people’s first hunches and the responses they gave after deliberation predominantly remained biased from start to finish. But in the rare cases in which participants did learn to correct themselves, they immediately managed to apply the solution strategy and gave a correct hunch on the subsequent problems. We discuss critical methodological and theoretical implications.
