Objective stepwise Bayes weights in survey sampling

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Jeremy Strief and Glen Meeden1


Although weights are widely used in survey sampling their ultimate justification from the design perspective is often problematical. Here we will argue for a stepwise Bayes justification for weights that does not depend explicitly on the sampling design. This approach will make use of the standard kind of information present in auxiliary variables however it will not assume a model relating the auxiliary variables to the characteristic of interest. The resulting weight for a unit in the sample can be given the usual interpretation as the number of units in the population which it represents.

Key Words

Sample survey; Weights; Bayesian inference.

Table of content

1 Introduction

2 The Polya posterior

3 The constrained Polya posterior

4 Constrained Polya posterior weights

5 The weighted Dirichlet posterior

6 Weights and Horvitz-Thompson

7 Examples

8 Final remarks






1Jeremy Strief, Principal Statistician, Medtronic Energy and Component Center, Brooklyn Center, MN 55430. E-mail : jstrief@gmail.com; Glen Meeden, School of Statistics, University of Minnesota, Minneapolis, MN 55455. E-mail : glen@stat.umn.edu.

Date modified: