Experimental evaluation of random incentive system under ambiguity

The random incentive system (RIS) is a standard incentive scheme used to elicit preferences in economic experiments. However, it has been speculated that RIS may not be incentive compatible when participants are concerned about ambiguity, i.e., that the choices observed under RIS do not reflect the underlying preferences. To examine the performance of RIS under ambiguity, we conducted three experiments online and in a laboratory. The results of the experiments suggest that RIS is incentive compatible. We argue that presenting choice situations in isolation may improve the incentive compatibility of RIS. We also argue that using RIS, together with an experimental guideline called Prince, may reduce the observed ambiguity aversion.