|
Activity Number:
|
99
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #303608 |
|
Title:
|
Model-Based Approach to Small Area Estimation of Disability Count and Rate Using Data from the 2006 Participation and Activity Limitation Survey
|
|
Author(s):
|
Valerie Bizier*+
|
|
Companies:
|
Statistics Canada
|
|
Address:
|
100 Tunney's Pasture Driveway, Ottawa, ON, K1Y1N9, Canada
|
|
Keywords:
|
small area estimation ; disability counts and rates ; Log linear unmatched model ; Gibbs sampling method
|
|
Abstract:
|
This paper presents the small area model considered to produce estimates of the number of persons with disabilities and disability rates for health regions and selected municipalities using the 2006 Participation and activity limitation survey data. The paper describes the transformations applied to direct estimators and to the variances associated with these estimators in order to meet certain fundamental criteria. The log linear unmatched model to which the hierarchical Bayes (HB) approach was applied by relying on the Gibbs sampling method, and the results from the latter model are presented. Lastly, the paper presents the data sampling methodology used to ensure that the final statistics take into account province level results.
|
