Online Program

Use and quality of auxiliary data informing sample designs for population surveys of the disabled

*Julian Whiting, Australian Bureau of Statistics 
Simon Strickland, Australian Bureau of Statistics 
Lyndon Ang, Australian Bureau of Statistics 
Kelly Chiu, Australian Bureau of Statistics 

Keywords: sample design, auxiliary data, disabled population

The Australian Bureau of Statistics’ Survey of Disability, Aging and Carers (SDAC) plays a major role informing Government health and welfare policy for the severely and profoundly disabled. It supplies key estimates on both the prevalence and characteristics of this important subpopulation which comprises around 6% of the Australian population.

The survey has traditionally adopted a two-stage area-based design with equal probability sampling. The inclusion of questions regarding disability status in the five-yearly population Census raised hopes of improving design efficiency by using this information to sample more heavily from areas with a higher density of disabled persons.

This paper presents methods and the results of an investigation into the potential efficiency gains from adopting this sample design approach for the 2012 SDAC. The study concludes the potential efficiency gains can be quickly eroded when the density information becomes degraded due to factors including population migration.

The second part of this paper considers the SDAC sample design in the context of the Bureau's future direction of selecting household survey samples from an address list.