ADR UK Data First Evaluation Fellowship

直到最近，在罪犯被刑事司法系统移送时例行收集的大量行政数据还无法用于研究。相反，我们的理解在很大程度上局限于对旅程中特定点的静态洞察。英国司法部ADR联动项目改变了这一局面，允许罪犯在刑事司法系统的各个阶段被追踪。这为认识到刑事司法系统不同部分之间存在的复杂相互联系的前沿研究开辟了巨大的潜力。例如，这可以帮助我们理解为什么一些被法院判刑的人在刑期结束后很快（而且反复地）回来，而另一些罪犯却再也没有出现过。也可以检查更复杂的刑事司法历史的影响，包括通过系统的多次过渡，以及特定干预措施对特定类型个人的影响。但是，新一波关联数据的复杂性和规模需要新的工作方法和对新的分析技术的理解。该奖学金是一个无与伦比的机会，可以直接与司法部分析师一起工作，以实现这些关联数据的全部潜力。通过合作，我将确定一些明确定义的研究问题，这些问题符合司法部的优先事项，可以用这些数据来解决。特别是，将利用将不同刑事司法阶段的资料联系起来所提供的机会。具体的问题将由我自己的学术理解通过刑事司法系统建立了超过15年的经验犯罪学家个人旅程的指导。它们还将适当地反映关联数据源中固有的结构复杂性，包括正确处理上下文的作用（在特定法院和/或监狱中处理的影响）以及先前经验影响后续经验的事实。研究问题将在一个互动讲习班上与司法部分析人员协商后加以完善。这将通过使用现有数据进行一些初步的“概念验证”数据分析练习来实现。这里的重点将是为进一步的讨论提供快速的证据基础，而不是选择技术上最复杂的分析解决方案。这些快速的数据深入将有助于突出具体的数据挑战，澄清中心研究问题，并促进进一步的讨论和问题的发展。然后，最有希望的问题将被纳入全面的实证项目，并得到更可靠的统计分析方法的支持，这些方法适当地反映了数据中出现的模式。奖学金的关键是确保为未来的研究留下遗产。为了实现这一目标，除了在学术和政策渠道上更标准地发表关键发现外，所有工作都将写在数据分析工作表中。这些工作表直接链接到原始数据，并将运行完成数据分析所需的计算代码，同时还包括解释性文本和可发布的输出。将数据处理、分析和报告的所有元素保持在同一工作表中，将确保经验工作的完全可复制性，同时还允许新用户快速调整代码和/或文本以生成新报告。我还将在相关的地方为司法部分析师举办讲习班。

Until recently, the large amounts of administrative data routinely collected about offenders as they are moved through the Criminal Justice System have been inaccessible to research. Instead, our understanding has largely been restricted to static insights into particular points in the journey. The Ministry of Justice ADR UK linkage project has transformed this picture, allowing offenders to be tracked across all stages of the Criminal Justice System. This opens up vast potentials for cutting edge research that recognises the complex interconnections that exist between different parts of the Criminal Justice System. For example, this could help us to understand how and why some of those people that are sentenced by the courts return quickly (and repeatedly) following the completion of their sentence, whilst other offenders are never seen again. The effects of more complex criminal justice histories including multiple transitions back through the system can also be examined as well as the impacts of particular interventions on particular types of individual. But the complexity and scale of this new wave of linked data necessitates new working approaches and understanding of new analytic techniques. The fellowship is an unrivalled opportunity to work directly alongside Ministry of Justice analysts to realise the full potential of this linked data. Working collaboratively, I will identify a number of clearly defined research questions that meet Ministry of Justice priorities and can be addressed with this data. In particular, the opportunities afforded by linking information from across different Criminal Justice stages will be exploited. The specific questions will be guided by my own academic understanding of individuals journeys through the Criminal Justice System built up over more than 15 years as an empirical criminologist. They will also appropriately reflect the structural complexities inherent in the linked data sources including correctly engaging with the role of context (the effect of being dealt with in a specific court and/or prison) and the fact that prior experiences shape subsequent ones. Research questions will be refined in consultation with Ministry of Justice analysts in an interactive workshop. This will be informed by some initial 'proof of concept' data analysis exercises with the available data. Here the emphasis will be on providing a rapid evidence base for further discussion rather than on selecting the most technically sophisticated analysis solutions. These rapid data deep-dives will help to highlight specific data challenges, clarify the central research question, and facilitate further discussion and question development. The most promising questions would then be worked up into full-scale empirical projects supported by more statistically robust analytic approaches that appropriately reflect the patterns that emerge in the data. Crucial to the fellowship is ensuring a legacy for future research. To achieve this, in addition to the more standard publication of key findings in academic and policy outlets, all work will be written up within data analysis worksheets. These worksheets link directly to the raw data and will run the computational code required to complete the data analysis, whilst also including explanatory text and publishable outputs. Keeping all elements of the data processing, analysis, and reporting within the same worksheet will ensure fully replicability of the empirical work, whilst also allowing new users to quickly adapt the code and/ or text to generate new reports. I will also run workshops for Ministry of Justice analysts where relevant.