A decision maker is presented with a theory from a self-proclaimed expert about the probability of occurrence of certain events. The decision maker faces the possibility that the expert is completely ignorant about the data generating process and so she’s interested in mechanisms that allow her to screen informed experts from uninformed ones. The decision maker needs to control for type I error, however, since she’s also uncertain about the true stochastic process, this gives room for uninformed experts to make strategic forecasts and ignorantly pass tests and profit from contracts. We present an original multiple expert model where a contract achieves screening of informed and uninformed experts by means of pitting experts’ predictions against each other. Additionally, we present a theoretical review of the main findings in two branches of literature that attempt to solve the expert screening problem. Namely models about testing experts and models in contract theory that pursue screening of experts.
