Managing Registered Sex Offenders

Tabled: 28 August 2019

Appendix B. Data quality checks

Cross-referencing SOR data with LEAP data

At the time of this analysis, there were 8 004 RSOs in the SOR database. These were cross-referenced with the LEAP database and examined for consistency.

We found that:

  • 7 219 records, or 90.2 per cent, had matching dates of birth and names
  • 427 records, or 5.3 per cent, had matching dates of birth but variations to the last name
  • 346 records, or 4.3 per cent, had the same or recognisably similar names but dates of birth were not matching
  • 12 records, or 0.2 per cent, were not found in LEAP at all.

We worked through these discrepancies with Victoria Police, but agree that the variations in names and dates of birth do not have an impact on RSO management in practice because the records are linked by unique identifiers.

Victoria Police could also explain the 12 files that were not in LEAP as being either duplicate SOR database entries or SOR database entries with multiple unique identifiers in LEAP. This can occur when RSOs have been entered on the SOR database with the unique identifier that was current on LEAP at date of entry and LEAP has subsequently been altered to reflect a new primary unique identifier. This usually occurs where what were thought to be two separate people are identified as the same individual. Entering the unique identifier recorded on the SOR database into LEAP automatically substitutes and redirects to the other, more current one and as such does not have an effect in practice. The SOR unit advises that they have now corrected the SOR data to reflect the most current unique identifier on LEAP.

Cross-referencing LEAP data with the SOR database

All RSOs must be given an RSO 'flag' on their corresponding record in LEAP. We cross-referenced these two datasets and assumed that any person flagged as an RSO in LEAP would have their details recorded in SOR.

There were 211 RSO flags in LEAP at the time of the analysis that were not found in the SOR dataset that we had received. Of these, 181 records can be explained as a timing difference between the SOR and LEAP data extracts—they were recorded in LEAP after the SOR dataset was extracted. The remaining 30 records, or 14 per cent, were not found in the SOR database. Victoria Police checked these records and the reasons for the discrepancies included:

  • duplicate unique identifiers in LEAP
  • test records in LEAP created by Victoria Police IT staff responsible for LEAP quality assurance which cannot be deleted
  • a delay in creating the SOR record as the LEAP flag is entered immediately but the SOR record may require paperwork follow up (Victoria Police confirmed the nine examples this applied to are all now in SOR).

There were 12 instances where the RSO flag was raised in relation to RSOs who are not registered in Victoria but registered in another jurisdiction. At some point, Victoria Police believed they were moving into Victoria and would be registered which led to the addition of the flag. This ultimately did not occur, and Victoria Police has since confirmed these flags have been removed.

Back to Top