Initial results from the 2002 NATSISS were released in the National Aboriginal and Torres Strait Islander Social Survey 2002 (ABS 2004c) and a series of web-based State and Territory spreadsheets (www.abs.gov.au). These releases covered topics as diverse as family and culture, health, disability, education, work, income and housing, law and justice, information technology and transport.

In response to the expectations of policymakers and researchers, the ABS has undertaken a range of initiatives to provide users not only with relevant statistics, but also with associated information about the quality of those statistics. In particular, ABS has calculated precise estimates of sampling error for 2002 NATSISS data using a grouped jack-knife methodology based on partitioning the primary sample into 250 cross-sectional groups (Wolter 1985). Estimates of relative standard errors have been released for all data shown in the main summary of findings report and in the web-based output at State and Territory level. The 250 replicate weights used in the jack-knife methodology are available on the CURF. Further, the results of tests of statistical significance are shown in selected tables in the main report and web-based spreadsheets, to inform users of the statistical significance of differences between estimates, for example in the measures of change over time or between the Indigenous and non-Indigenous populations.

Since the Indigenous population has a younger age structure than the non-Indigenous population, use of age-standardisation is sometimes appropriate when comparing estimates from the two populations. In the national report, age-standardisation was used for health, employment, law and justice, and information technology items, as these data items were judged to be associated with age in ways that warranted this statistical technique.

Responding to user demand for unit record data in order to facilitate more flexible and detailed analysis of survey files, the ABS developed the RADL in 2004. This facility allows users to access CURFs over the web. By maintaining the data file within the ABS computer environment and monitoring its use by researchers, the ABS is able to provide a more expanded data file than would be possible if the CURF was released on CD-ROM. While universities have access to CURFs via the RADL under an arrangement between the ABS and the Australian Vice-Chancellors’ Committee, other agencies and researchers can purchase a licence to access CURFS via the RADL (i.e. over the web).

The 2002 NATSISS CURF contains a higher degree of disaggregation than the 1994 survey CURF. In particular, the 2002 NATSISS CURF includes geographic items by State/Territory as well as by remoteness area. For more information, see the National Aboriginal and Torres Strait Islander Social Survey CURF Technical Paper (ABS 2005b).

The strength of the 2002 NATSISS data set is that the broad range of socioeconomic variables enables the exploration of associations among different outcomes, including analysis of multiple disadvantage. To this end, the ABS is engaging in a number of collaborative research projects to examine different topics, including self-assessed health, victimisation and cultural issues.