30th September 2016 Delivered in Leeds by Lancaster University
Over forty people recently attended an N8 PRP workshop as part of the training & learning strand of the project. ‘Data Analysis, Management and Crime Mapping’ was part of a series of workshops aimed at developing research training and learning among police and partner agencies.
This was a hugely successful event which attracted delegates from police forces, academia, and other policing agencies.
This event brought together practitioners and academics to discuss best practice and innovative approaches to analysing and managing data, as well as how data mapping techniques have been used to better understand crime patterns.
Topics covered during the event included:
- Predictive modelling techniques and crime ‘hotspots’.
- Measurement of crime data and its impact on our understanding of crime.
- Theorising the causes of crime.
- Agent-Based Modelling.
- Practical application of research.
The workshop was started with a brief overview of the N8 PRP Data Analytics strand with plans for the future and information on how data analysts might further get involved with the project. If you would like more information on the Data Analytics strand then please contact Fiona McLaughlin.
This was followed by a series of workshops:
The way we count has a significant impact on the data we collect, the methods we use to analyse it, the findings produced and thus what we know about the social world and how we design our research, policy and practice interventions. What or who is included or excluded, the boundary of our definitions, and the theory of change we seek to test and develop (whether implicit or explicit) are fundamental parts of the process and require critical examination.
Using a case study on violent crime, Dr Towers explored what happened when a statistical solution to a substantive problem was challenged and a new methodology developed for estimating violent crime using data from the Crime Survey for England and Wales. This new methodology takes full account of high frequency victimisation, and does not ‘cap’ the number of repeat victimisations counted in the way the official method does. The findings resulting from this new methodology suggest a reversal of the official trend and increasing rates of domestic violent crime, violent crime against women and ultimately in violent crime in England and Wales. This was a clear account of the importance of understanding how data has been collected and presented to inform new methodologies for analysing crime contextualised within a developing Theory of Change.
Dr Towers further supported a recurrent theme across all the talks – the need to theorise the causes of crime in order to prevent crime, thus the relationship between theorisation, the production and use of ‘evidence’ and theory-testing are crucial for effective evidence based policing.
This presentation is available to download here
Les Humphreys, Lancaster University – ‘Using the PNC as a means to measure and analyse organised crime’
Following on, Dr Les Humphreys also unpicked the process of measuring and analysing data in relation to organised crime, to provide results which might then effectively inform strategic leaders when considering particular organised crime priorities. Dr Humphreys used some of his recent work to highlight some of the key factors to consider when sampling and filtering data for analysis, highlighting this importance of first effectively defining your crime area and understanding the possible lack of standardisation in data collection. The presentation was a detailed review of a methodology to provide valuable empirical evidence to improve the evidence base for policy and tactical interventions, in this case in relation to organised crime but that might be translated to other areas of crime too.
This presentation is available for download here
Ken Pease, UCL & David Oldroyd, formerly West Yorkshire Police – ‘Growing Your Own: The Experience of Local Approaches to Predictive Patrolling’
Predictive modelling techniques have been utilised by police constabularies to tackle crime ‘hotspots’ and predict and then prevent crime. This presentation explored how police crime data can be used to estimate the probability that crime may be more likely to occur in a specific location at particular times and how this has been a useful basis for optimising resource allocation. In addition, how the analysis and management of data has been shown to be effective in the detection and prevention of specific crimes and the likelihood of repeat victimisation. The session looked at the example of Operation Optimal, which David headed. The obstacles to its implementation were identified. The
success that Optimal achieved was only made possible by a process of continual review and revision of practice which calls into question the value of research designs which take an initiative as complete at the point of its launch.
This presentation is available for download here
Dr Nick Malleson and Dr Andy Evans also looked at predictive modelling techniques through ‘Agent-Based Modelling’. This looked at a computer simulation, developed by Dr Malleson, that can be used to simulate the behaviour of virtual people in a realistic urban environment (a virtual space with houses, roads, communities, railways etc). Within this virtual environment, components of can be tweaked to see how the crime system responds and thus can be used to determine what effects it could have on burglary rates and in forecasting expected crime areas or ‘hot-spots’. Agents within the model are developed using existing theory and this is continuously tested to provide statistical validation that these models do work. This is a forward facing, technological approach to
predictive policing and more work is also being done to see how this might be used beyond burglary to predict offender behaviour and to explore how geo-located social-media data can be used to further inform predictive models.
This presentation is available online here
Scott Keay & Russell Clark, Lancashire Constabulary – ‘Appliance of science: the practical application of criminological research to the workplace’
Both Scott and Russell work on the community safety partnership panel at Lancashire Constabulary and provided the audience with examples of work already going on at Lancashire, stressing the importance of using datasets from a variety of agencies, understanding the interpretation and presentation of that data, and championing collaboration between academia and policing to help improve the understanding of harm and wider aspects of crime. The presentation looked specifically at geo-spacial analysis and some of the pitfalls of using maps as a predictive tool, arguing that maps alone should not inform practice and should be framed with theory as part of a wider evidence based approach to be effectively translated into practical action. This convincingly highlighted the need for a more holistic approach to underpin effective decision making.
Lancashire are using a variety of methods to encourage evidence based practice (EBP):
- EBP Cafes – Targeting front line officers to inform and engage to promote leadership from the bottom up.
- Academic Review Exercise – Each member of the team produce an academic review each month. This provides continuous professional development and ensures knowledge of up to date research
- 8 Step Approach – Using eight points that are essential for successful initiatives: Community focused (with local commitment); Theory based (based on sound research); Specific purpose (clear objectives); Innovative / creative; Evaluation criteria set out prior to implementation; Evaluation conducted; Sufficient resources to deliver; Exit strategy.
- Research Framing – Investing in academia and engaging with students and volunteers to produce lit reviews and work with analysts.
The presentation provided a case study which looked at ‘An Evidence Based Approach to Predicting Risk Areas for Domestic Burglary’ . This showcased the effectiveness of the approaches that had been outlined during the session, showing the practical implementation of existing theory to increase the forces predictive power. The Triple-T strategy (Sherman, 2013) was suggested to be a strong approach, using Targetting, Testing and Tracking as a basis for this model.
This presentation is available for download here
For further information please contact Russel Clark
Context and Spatial Nuance Inside a Neighborhood’s Drug Hotspot: Implications for the Crime–Health Nexus – Curtis et al.
During roundtable discussion delegates were asked to think about their top three priorities for their own future training needs, an evaluation of the topics suggested will inform a series of further workshops over the coming months. A timetable of forthcoming events with details of how to register will be published in the next newsletter.
In the meantime if you have further feedback on the training and learning strand, including suggestions for training and learning events get in touch with the acting Training and Learning strand lead Dr Jude Towers at Lancaster University: email@example.com
The Training and Learning strand will also be launching a programme of research to identify current training and learning practices in the N8 PRP and ways in which these can be supported and developed by the N8 PRP project – further details in the next newsletter…
“Through our programme of training and learning we are seeking to enhance the research skills and use of evidence among policing partners, to secure research impact and to maximise the practical benefits to policing innovation and exploitation of data.”
Professor Adam Crawford,
N8 PRP Director