DATA QUALITY STATEMENTS

The purpose of these statements is to communicate data quality information to users. The ABS aims to produce high quality data from the Census. Quality management of the range of different processes involved in the Census aims to minimise errors as much as possible. Therefore, it is important that users are alerted to any known quality issues and provided with a measure of any remaining errors so data can be used in an informed way.

Data Quality Statements are available for all standard data items from the Census. The statements provide information on:

• coding discrepancy rates where applicable
• processes and edits applied to the data, and
• imputation rates.

General quality information about 2016 Census data will be provided with each data release. More extensive data quality information will be added as it becomes available.
PERTURBATION

All tables of Census data are subject to perturbation to protect the confidentiality of individuals. This is in accordance with the Census and Statistics Act 1905 which stipulates that no results will be released in a way which would likely enable particular individuals or households to be identified.

Perturbation is a technique which has been developed to randomly adjust count values. When the technique is applied, all counts and totals are slightly adjusted to prevent any identifiable data being exposed. These adjustments result in small introduced random errors. However, the information value of the table as a whole is not impaired. This technique allows very large tables to be produced even though they contain numbers of very small counts.

Perturbation is applied in a controlled manner so the same counts of information are consistent between tables. In addition the tables at different geographic levels are adjusted independently. In general, users should construct a table with the information they require so perturbation is applied to a count only once; it is not recommended to sum across perturbed counts to derive the information you require.

No reliance should be placed on small cells as they are impacted by perturbation, respondent and processing errors.

Perturbation can be a source of frustration to users, particularly those with a lesser degree of familiarity with Census data, as it can result in inconsistencies in the data with rows or columns not adding to displayed totals. Most tables reporting basic statistics at an elementary level will not show discrepancies due to random perturbation. However, as the degree of sophistication of tables increases, the need for random perturbation remains and it will continue to be used in the release of data for the 2016 Census.