Providing expertise to the Australian Bureau of Statistics to meet its vision of unleashing the power of statistics for a better Australia

Reliable statistics are critical to the effective function of government, and as society changes, so does the need to adapt collection and analytical statistical methods.

A team of University of Wollongong researchers from the National Institute for Applied Statistics Research (NIASRA) is partnering with the Australian Bureau of Statistics (ABS) to provide an evidence-base for innovative statistical methods.

The partnership was established in 2005 to encourage research into new and innovative methods of collecting, analysing and using statistical information relevant to the ABS, and included funding for a Professor of Statistical Methodology for 2006-2014.

The group at NIASRA has expertise in survey design and analysis, complex data analysis and estimation methods, and privacy and confidentiality analysis, among a number of other areas of interest to the ABS.

According to NIASRA’s Distinguished Professor Noel Cressie, producing reliable estimates for small geographic areas from sample surveys is challenging for agencies like the ABS that report on the social and economic conditions in the country.

“This project is developing and testing new methods to produce estimates for small area where the surveys are conducted repeatedly across time, by using spatial and temporal relationships, as well as other variables,” he said.

The team is also working together on a project that is considering the selection variables for calculating weights used in complex multi-objective household surveys. The team will propose a final weighting strategy for the 2016-17 personal safety survey.

The ABS has supported NIASRA research through five Australian Research Council Linkage grants. This funding is specifically designed to connect research and industry.

As well as providing opportunities to deepen the partnership between the two organisations, the grants created opportunities for post-doctoral research positions and PhD students, and helped grow NIASRA’s international collaborations.

“The common interest in advanced statistical methodology should ensure continuing collaboration with the ABS through joint research and funded projects,” NIASRA Director Professor David Steel said.