Some differences in revealed behaviour under different inquiry methods

The experiment presented in this paper has two aims, both methodological. First, we want to check for the role of what we may call (after Carpenter et al., 2006) the they came to play effect. Second, we want to test whether the lab outcomes are confirmed by a questionnaire on a hypothetical similar scenario. In order to pursue our aims, we design an experiment made of four treatments: a lab-experiment with strategy method, a lab-experiment without strategy method, a questionnaire with strategy method and a questionnaire without strategy method. We may conclude that the lab results are definitively more reliable than the questionnaire ones only if you manage, in one way or the other, to get rid of the bias induced by the they came to play effect: a post-experiment questionnaire, containing explicit questions on the matter, may be a device.