Experimentally Testing Game Theory

摘要: Controlled economic experiments have the potential to provide direct tests of underlying assumptions game theory, rather than ambiguous typically provided by observational data. Thus, if a theory is not supported experimental data, we can eliminate confounding explanations like mis-specification situation, because situation constructed exactly match specification game. This allows us identify other, deeper, causes theory’s failure predict; further be designed tease apart competing for its failure, and suggest refinements which incorporate existing evidence. As result, much richer body that explain predict behavior in wide variety business settings.