Contigent Belief Updating

We study the effect of contingent thinking on belief updating. One possible explanation for biased beliefs is that agents distort the underlying data-generating process when updating their beliefs given a new piece of information. Engaging in contingent thinking – jointly assessing beliefs for all possible events without learning the realized one – might affect the agents’ understanding of the data-generating process, resulting in differences in belief updating. With a series of online experiments, we aim to uncover and decompose the effect of contingent thinking on belief distortions into two components: (1) hypothetical thinking (updating on a piece of not yet observed information) and (2) contrast reasoning (comparing multiple contingencies during the updating task). We propose to proceed in three consecutive steps. First, we plan to establish this effect in a neutral task and investigate possible mechanisms in terms of the data-generating process and individual characteristics. Second, we want to compare the impact in other domains, specifically in an ego-relevant task. Third, we intend to examine whether contingent belief updating translates into changes in actions and the role of commitment in beliefs.