Ulric B. and Evelyn L. Bray Social Sciences Seminar

Abstract: An analyst is tasked with estimating the parameter of a statistical model. They are given a performance score that they seek to maximize. The score may depend on the analyst's report and data generated by an independent experiment. We describe the information that can be elicited and apply our results to a variety of common statistical models. We then compare experiments based on the information they enable us to elicit and based on the incentive power they can induce. These orders are connected to, but different from, the Blackwell order. Our results shed light on how using data for incentive provision differs from using data for statistical inference.

Joint work with Yaron Azrieli, Christopher Chambers, and PJ Healy