Anmerkungen:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments November 10, 2015 erstellt
Beschreibung:
In many economic environments, people need to learn from systematic absences and non-occurrences. For example, news media differentially reports on events and corresponding non-events and social networks like Facebook selectively tailor their newsfeeds according to users' prior behavior. This paper studies people's belief formation rules in such selection contexts through a series of tightly structured experiments. Across various treatment variations, some subjects fully correct for selection, but many fail to take into account information they do not see and state beliefs that reflect exactly a full neglect benchmark. Follow-up treatments characterize the effects of environmental complexity on the presence of the updating bias and investigate the mechanism through which complexity operates. The results document that people's updating rule systematically depends on computational complexity and that complexity matters because it affects what people pay attention to: moderate complexity appears to induce cognitive busyness, which in turn distracts subjects from attending to the selection effect in the background of the process. Hence, most people can be debiased through a simple attentional nudge. However, exposing subjects to the conflicting beliefs of more rational participants is not sufficient to draw their attention to their mistaken reasoning