Umass Dept. of Biostatistics and Epidemiology Research
Improving Analysis Methods for Clustered Randomized Primary Prevention Intervention Trials
BioEpi :
Cluster Home
Contacts : Other Links


Current Topics
Dialogues/Comments
Outline
Introduction
  Exchangeability
Superpopulations
Single-stage Samp.
Two-stage Samp.
Estim. Domains
One Factor Study
Response Error
Mixed Models
Foundations of Stat.
Results
 Intranet
Bibliography

Estimating Domains
 

Inference for domains in single stage sampling: equal size domains
Synopsis:

We develop estimating equations for domain means based on a simple random without replacement sample when the domain sizes in the population are equal. In the population, domains are equivalent to clusters, or strata. They represent identifiable groups of units. This development closely parallels the development for a single stage sample given in c00ed27v1.doc and two stage sampling (c00ed67.doc). A general solution is given, but no explicit solution or examples are given.
  Contents: Introduction
The superpopulation framework: domain estimates under simple random sampling
Definition of the population and superpopulation
Models for the superpopulation
Expected value and variance of the superpopulation
Sampling, re-arranging and partitioning
Estimation
Estimating equations and estimates
  Print: 12 pages ; 9/22/00 ; c00ed68.doc
  Author: Ed Stanek ; Reviewer: none
  Related:  

Cluster Home
TOP : Outline :
University of Massachusetts at Amherst Copyright 2000 University of Massachusetts, Amherst.
This is a page at the University of Massachusetts Amherst Campus.
Produced and maintained in the Dept. of Biostatistics and Epidemiology at the School of Public Health and Health Sciences.
Send comments or questions about this web site to Webmaster : Oscar Loureiro at oscar@math.umass.edu
This page was last modified on October 17, 2003.