Scottish Crime Data Hackathon: Summary and reflections 

Aims and Overview 

The Scottish Crime Data Hackathon ran between 25-27 July 2023 at the Universities of Stirling and Edinburgh. It brought together 12 researchers across career stages and disciplines from criminology to biology, and from institutions across Scotland and England (both attending on campus and virtually) to produce and document a short piece of research using the Scottish Crime and Justice Survey (SCJS). 

The hackathon pilot projects were intended to advance understanding of contemporary crime and criminal justice topics through empirical research, as well as encouraging collaboration between criminologists and quantitative researchers in other disciplines. Projects were based on research themes identified by Scottish Government, specifically: 

  • Violence against women and girls 
  • Levels of violent crime victimisation, characteristics of victims 
  • Understanding experiences of crime by multiple forms of identity e.g. gender, ethnicity, disability 
  • Characteristics of people who are repeat victims (compared to one-off victims and/or non-victims) 
  • What characteristics discriminate between women who are victims of sex offences / domestic abuse and those who experience other types of crime? 

At the event we had presentations from representative from the Scottish Police Authority and Community Justice Scotland about how they use SCJS to inspire participants to come up with research questions, as well as presentations from Scottish Government’s Justice Analytical Services team, Ipsos and ScotCen about the survey design to help teams plan their analysis. 

 

The research projects 

  • Team One investigated the impact of personal and place-based characteristics on confidence in the police in Scotland using machine learning (Yunus Serhat Bicakci, Manikandan Soundararajan and Joe Schofield) 
  • Team Two explored whether gender-based violence (GBV) was more prevalent in rural or urban areas in Scotland, and whether GBV was more prevalent in areas of high deprivation (Trystan Washburne, Ben Moews, Oana Petcu and Ringmichon Keishing) 
  • Team Three provided an autoethnographic account of the Hackathon itself, whilst reflecting on confidence in the police in Scotland (Siria Capezzani, Sam Paplauskas, Jim Watson and Angus Bancroft) 

 Click on each of the links above to read the teams’ outputs. 

Reflections: Ben Matthews 

First, I would like to thank all the stakeholders who contributed their time and energy to the Hackathon – colleagues from Justice Analytical Services in Scottish Government, the Scottish Police Authority,  Community Justice Scotland, Ipsos and ScotCen all gave up their time support the event, for which we are extremely thankful. 

The hackathon certainly achieved its aims of bringing together a diverse group of researchers, and this was reflected in the participant feedback, with one participant reflecting after the Hackathon that “I got really good experience of working with people from very different disciplinary backgrounds”, and another that they had a “first exposure to new methods: machine learning, autoethnography”. 

However, there was also a sense that we asked a lot of participants to design and delivery a mini research project in just a few days. Combining nine sweeps of SCJS and navigating changes in variable names and definitions as the survey has evolved is a difficult thing to do, even for those who have worked with the survey before. Despite having expert help available in the form of Stuart from Justice Analytical Services, and Stuart having put together a harmonized dataset to for participants to work with, this proved challenging for participants.  

Mostly I am extremely proud of our participants and their willingness to throw themselves into the unknown. We asked a lot of participants in design and deliver research project in three days with collaborators they had never met before, and it can be very vulnerable to be in such an unfamiliar situation – either by virtue of analysing criminological data (for those from other disciplines) or working with data at all (for our qualitative criminologists). The quality of work they produced in such a short time is remarkable – there is something thought-provoking in each of these reports, in their conclusions and their reflections on the challenges they faced. For me, this makes the Scottish Crime Data Hackathon a resounding success. 

 

Stuart Napier (Justice Analytical Services, Scottish Government) 

Firstly, thanks to the SCCJR for involving me in the process and putting the SCJS forward to a new cast of people. It was fantastic to be able to witness external users of our data get to grips with it in real time which is a truly unique and valuable opportunity afforded by the event. I was very impressed by the participants’ attitudes to try something completely alien to them – especially given the time and resource constraints. Full credit to each of them for challenging themselves. 

The hackathon was doubtless a positive experience for the SCJS team in creating and improving links with users of our data. For me personally, one of the main takeaways was understanding the barriers that can prevent people from easily accessing and analysing the SCJS data. The event has inspired me to take forward work to broaden the audience of the SCJS and aid more people in using the vast and rich data produced by the survey. 

 

New Media, Surveillance and Technology

Associated People

Dr Ben Matthews

University of Stirling

Ringmichon Keishing

University of Stirling

Oana Petcu

University of Glasgow

Siria Capezzani

University of Stirling

Jim Watson

Edinburgh Napier University & Dundee University