|
Activity Number:
|
130
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, August 7, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
General Methodology
|
| Abstract - #305002 |
|
Title:
|
The Design and Analysis of Two-Phase Sampling
|
|
Author(s):
|
Alastair J. Scott*+ and Christopher J. Wild and Yannan Jiang
|
|
Companies:
|
The University of Auckland and The University of Auckland and The University of Auckland
|
|
Address:
|
36 Princes Street, Auckland, 1, New Zealand
|
|
Keywords:
|
two-phase sampling ; semiparametric efficiency ; auxiliary information ; biased sampling
|
|
Abstract:
|
In many health surveys, information about some characteristics is available for all units in the population; an example is the Auckland Collaborative Birthweight Study. Birthweights and many other characteristics are collected routinely for all mothers and babies in the target population. Other characteristics can be measured for a sample only. We want to build a regression model for birthweight as a function of the other characteristics. The task is to design an efficient sampling scheme and method for fitting the model, using all the information available. We will look at methods for computing efficient semiparametric estimates and alternatives based on standard survey techniques for using auxiliary information. These alternative methods are reasonably efficient and relatively robust and simple to implement. We also look at some aspects of design.
|
