Cognitive Economics in the Lab. The Power of Labels

It will be argued that the inclusion in the experimental design of contextual clues, such as “label” or “virtual simulations”, which make the laboratory task familiar, proved that problem solving in decision-making under uncertainty is environmentdependent. This supports the belief that the context-free experiment is not only an elusive goal, but also an inappropriate tool to investigate cognitive processes. To support this view, it will be presented an experiment on travel mode choice, in which subjects acquire ex-post information not only through their own experience but also with external information on actual travel times of all travel modes. In this setting, subjects exhibit a marked preference for cars, are inclined to travel mode stickiness, and update imperfectly their expectations on travel times. This implies that fully informed commuters choose travel modes by using some heuristic procedures that can lead to robust deviations from travel time minimization. It will be concluded that the use of virtual experiments combines the internal validity of controlled lab experiments with the external validity of field experiments.