|
Activity Number:
|
105
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 9, 2004 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #300849 |
|
Title:
|
Toward a Better Estimation of Working Residential Number Rate among the Undetermined: The Application of Survival Analysis in WRN Estimation
|
|
Author(s):
|
Lidan Luo*+ and Hiroaki Minato
|
|
Companies:
|
NORC, University of Chicago and NORC, University of Chicago
|
|
Address:
|
55 East Monroe St., Chicago, IL, 60603,
|
|
Keywords:
|
survival analysis ; unit response rate ; working residential number ; REACH 2010 Risk Factor Survey ; call history ; CASRO method
|
|
Abstract:
|
The AAPOR standard on response rate calculations permits some latitude in the assumptions made regarding the eligibility rate among cases for whom eligibility could not be determined. The CASRO method, for example, assumes the cases of unknown eligibility are eligible (WRNs) in the same proportion as cases for which WRN status can be determined. However, if the unknown cases are WRNs at a different rate, such a method might yield too inaccurate a response rate. The survival analysis approach described by Brick et al. (2002) takes advantage of the relationship between the level of difficulty (the number of calls) in reaching a household and the WRN rate. We demonstrate the method of survival analysis in estimating the WRN rate among the undetermined by simulations and a post-survey application to REACH Risk Factor Survey. Our results show that when more determined numbers are needed after the planned number of calls have been made, the survival analysis method seems always to give at least as good an estimate of the WRN rate among the undetermined as the traditional CASRO method does.
|
