An Introduction to the Local Pivotal Method and Variance Approximation Approaches

Document Type

Presentation Abstract

Presentation Date

4-11-2019

Abstract

Simple random sampling is a commonly used (and commonly taught) sampling method that is the extent of the knowledge of sampling theory for many people. Are there better ways of selecting a sample? If so, in what instances is one sampling design better than another? What does it even mean for one sampling design to be better than another? These questions will be explored through the introduction of some basic sampling designs, and through the definition of those designs more interesting questions about the way in which such complicated sampling designs can be carried out will be answered. Splitting methods will be introduced as a means of selecting unequal inclusion probability random samples. Two special splitting methods are the pivotal method and local pivotal method, the latter of which incorporates auxiliary information into the way that a sample is selected. One difficulty with these methods is that the variance of estimates can be a problem to calculate, so some of the current work being done on estimation of variance will be presented.

Additional Details

Thursday, April 11, 2019 3:00 pm in Math 103
Refreshments at 4:00 p.m. in Math Lounge 109

This document is currently not available here.

Share

COinS