Inferential models: A framework for prior-free posterior probabilistic inference

TL;DR

This paper introduces inferential models (IMs), a novel framework for prior-free probabilistic inference that guarantees data-dependent uncertainty measures with long-run frequency calibration, addressing limitations of previous methods.

Contribution

The paper develops a new IM framework that provides calibrated, prior-free probabilistic inference using auxiliary variables and random sets, with an optimality theory to resolve non-uniqueness.

Findings

IMs achieve frequency calibration under mild conditions

The framework provides data-dependent probabilistic measures

Examples demonstrate the effectiveness of the approach

Abstract

Posterior probabilistic statistical inference without priors is an important but so far elusive goal. Fisher's fiducial inference, Dempster-Shafer theory of belief functions, and Bayesian inference with default priors are attempts to achieve this goal but, to date, none has given a completely satisfactory picture. This paper presents a new framework for probabilistic inference, based on inferential models (IMs), which not only provides data-dependent probabilistic measures of uncertainty about the unknown parameter, but does so with an automatic long-run frequency calibration property. The key to this new approach is the identification of an unobservable auxiliary variable associated with observable data and unknown parameter, and the prediction of this auxiliary variable with a random set before conditioning on data. Here we present a three-step IM construction, and prove a frequency-calibration property of the IM's belief function under mild conditions. A corresponding optimality theory is developed, which helps to resolve the non-uniqueness issue. Several examples are presented to illustrate this new approach.