|
Activity Number:
|
380
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #304270 |
|
Title:
|
Frequency Matching and Balance in Case-Control Studies: Quantifying the Efficiency Loss of Robust but Suboptimal Design Strategies
|
|
Author(s):
|
Jennifer L. Wilcock*+ and Alan J. Lee
|
|
Companies:
|
University of Auckland and University of Auckland
|
|
Address:
|
Department of Statistics, Auckland, International, 1142, New Zealand
|
|
Keywords:
|
case-control study ; design ; optimal design ; two-stage ; frequency matching
|
|
Abstract:
|
Frequency matching and using two-stages of sampling are both well-established strategies aimed at improving the efficiency of case-control studies. For design, generally robust approaches are typically used but little is known about the optimal design of these studies. We systematically examine the efficiency loss when using robust traditional frequency matched and balanced two-stage designs. By adapting a semiparametric ML method developed for the analysis of two-stage case-control studies (Scott & Wild, Biometrika, 1997), an expression for the asymptotic variance of both designs has been derived and used to find optimal designs for studies with arbitrary number of strata. For frequency matching, the asymptotic efficiency of optimal and traditional matched designs are compared. For two-stage designs equivalent comparisons are made between optimal and balanced second stage designs.
|
