ABSTRACT

Police data and survey research provide different bases to inform research on crime and delinquency. We argue that linking police data on local crime incidences to criminological surveys allows for new insights on the role of residential and school contexts for juvenile delinquency and violence. We describe the challenges and solutions of combining these data sources in a collaboration between the state police of North Rhine-Westphalia – Germany’s most populous state – and social scientists from a major German university. In this academic-practitioner partnership, data from a four-wave longitudinal study of more than 3800 students were linked to spatially aggregated data from the police crime statistics for the years 2013–2016. We discuss how the simulation of nearby addresses can serve as a tool for anonymized data linkage, how knowledge of the local data collection practices is crucial to evaluate the geocoding accuracy of address-level crime data, and how sensitivity and implication analyses can help to reduce uncertainties at the analysis stage. We also give recommendations for future research and data collection practices.

摘要

警方数据和调查研究为犯罪和犯罪研究提供了不同的基础。我们认为，将当地犯罪发生率的警方数据与犯罪学调查联系起来，可以对住宅和学校环境对青少年犯罪和暴力的作用产生新的见解。我们描述了在德国人口最多的州北莱茵-威斯特法伦州警察与德国一所主要大学的社会科学家之间的合作中结合这些数据源的挑战和解决方案。在这个学术-从业者合作中，来自 3800 多名学生的四波纵向研究的数据与 2013-2016 年警察犯罪统计数据的空间聚合数据相关联。我们讨论附近地址的模拟如何作为匿名数据链接的工具，了解当地数据收集实践对于评估地址级犯罪数据的地理编码准确性如何至关重要，以及敏感性和影响分析如何有助于减少分析阶段的不确定性。我们还为未来的研究和数据收集实践提供建议。