|
Activity Number:
|
117
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Government Statistics
|
| Abstract - #307111 |
|
Title:
|
Using Sensitivity Analysis To Manage Uncertain Matching in Multiple Systems Estimation
|
|
Author(s):
|
Patrick Ball*+ and Scott Weikart and Rapheal Kaplan and Romesh Silva and Jeff Klingner
|
|
Companies:
|
Human Rights Data Analysis Group and Human Rights Data Analysis Group and Human Rights Data Analysis Group and Human Rights Data Analysis Group and Human Rights Data Analysis Group
|
|
Address:
|
480 California Ave., Palo Alto, CA, 94306,
|
|
Keywords:
|
multiple systems estimation ; record linkage
|
|
Abstract:
|
Multiple Systems Estimation (MSE) assumes that the identification of duplicates within a system and the record-linkage (matching) process between systems is error-free. When analysts work with data collected in unsettled situations, conflict zones and least developed countries, biographical information about individuals is often missing or incorrect. As a result, the matching decisions are subject to substantial uncertainty. Consequently, we developed a parameterized record-linkage process which generates multiple sets of matched data by varying the criteria upon which record-linkage decisions are made. We then calculate MSE estimates for all sets of the matched data and assess the distributional spread of the resulting estimates. Examples will be presented using mortality data from Timor-Leste and Sierra Leone.
|
