
Chapter 7
Marginal Inferential Models
Portions of the material in this chapter are from R. Martin and C. Liu, “Marginal
inferential models: prior-free probabilistic inference on interest parameters,” Journal
of the American Statistical Association, to appear, 2015, reprinted with permission
by the American Statistical Association, www.amstat.org.
7.1 Introduction
In statistical inference problems, it is often the case that only some component or,
more generally, some feature of the parameter θ is of interest. For example, in lin-
ear regression, with θ = (β,σ
2
), often only the vector β of slope coefficients is of
interest, even though the error variance ...