The Crime Statistics Agency has undertaken a quality assurance review of Aboriginal and Torres Strait Islander self-identification in our recorded crime dataset, as recorded by Victoria Police. These data have high proportions of missing or unknown values in their raw form and are not of sufficient quality for general use.
Given the high level of demand for statistics relating to Aboriginal and Torres Strait Islander people’s involvement in recorded crime, the agency has been exploring different options for calculating a person's Aboriginal or Torres Strait Islander status, to try and potentially 'improve' the quality of these statistics to a level where they are fit for purpose. These techniques generally involve using the values that are visible in the data, to fill in the gaps where values are unknown. Three were specifically proposed:
- an ‘ever identified’ counting rule;
- a ‘most frequent’ counting rule; and
- a ‘most recent’ counting rule.
As the key principle is that everyone represented in these data should be able to self-identify as to whether they are of Aboriginal or Torres Strait Islander origin, it was crucial that the ethics of these methods and implications were transparent and acceptable to community.
To determine which method would be most acceptable, the Crime Statistics Agency released a consultation paper in June to seek public comment and stakeholder views about which - if any - of these methods for calculating Aboriginal and Torres Strait Islander status in the recorded crime dataset should be used in producing our official recorded crime statistics for Victoria. You can find the original consultation paper which outlines the three different options here. The agency also consulted directly with a range of community representatives to seek views on the proposed methodologies. Thank you to all who provided their views.
The Crime Statistics Agency was keen to not make important decisions about how to represent Aboriginal and Torres Strait Islander people in statistics in isolation and to ensure that the views of community were reflected and the final statistical results are acceptable.
Overall, feedback received by the agency indicated that the ‘most frequent’ counting rule was the favoured methodology. Respondents also noted the desire for the raw data to be improved at the point of data collection, and for the agency to review the methodology used as data quality changed in the future. We will continue to monitor the quality of these data and consider our approach as necessary.
As a result of implementing this methodology, Aboriginal and Torres Strait Islander offender statistics will be included for the first time by the CSA in the upcoming 2015-16 annual report. This report will be released on the CSA website on 29 September 2016. Should you require any additional information relating to Aboriginal and Torres Strait Islander recorded crime statistics for Victoria, please contact us at firstname.lastname@example.org (External link) to discuss your requirements.