Data linkage

1.23 The team has successfully linked prisoner home address information with 2016 Scottish Index of Multiple Deprivation (SIMD) data to add insight on arrivals by deprivation. This analysis shows that people from more deprived areas are more likely to end up in prison than those from less-deprived areas. There are known quality issues with home address data, such as changes to address not being reported or recorded, or incomplete or incorrect postcode information, but the team accounts for this by excluding records with partial information and prisoners of no fixed abode. The technical manual is clear about the approach used, but it could explain what percentage of records cannot be matched, to give an indication of confidence in statistics produced from the linked data.

1.24 The team’s plans to link prison population data with data on a range of other topic areas, including education, health and drug use, show ambition to maximise the value of the statistics. It supports our vision of a statistical system that makes data linkage the norm. JAS is working with the Scottish Government Data Sharing and Linkage team on this. The statistics team is approaching this work in a methodical and rigorous way, by mapping out which data sources are available and which questions are being asked. The datasets are expected to be linked and made available through a safe haven, managed by Research Data Scotland. These datasets will help academics and researchers explore issues related to the prison population that can only be answered through sharing and linking data, for example, on prisoners’ educational background, health background, and drug use history. To demonstrate transparency around data linkage, it might be helpful to publish a summary of data linkage plans and update users on progress.