Ocena zagrożenia przestępczością na małych obszarach - problemy metodyczne
Assessing the risk of crime in small areas - methodological problems

Summary/Abstract: The objective of this article is to evaluate the viability of employing a selected set of statistical measures for comparing small geographical areas in terms of crime risk. The study is based on data on recorded crime in selected counties in Poland over the period 2020-2024. It is demonstrated that substantial, one-off fluctuations in crime rates, frequently observed at the level of small territorial units, may lead to erroneous conclusions when analyses rely solely on data from a single year or on average values. To address this issue, the article proposes the use of the median crime rate as an alternative comparative measure in situations where deviations from long-term trends occur, and discusses both the advantages and limitations of this approach. Spatial differentiation in crime across Poland is examined using methods of spatial statistics, including Moran's global statistic, while clusters of counties with similar levels of crime risk are identified using Ward's hierarchical method. Furthermore, a multidimensional taxonomic measure for assessing crime risk is proposed, based exclusively on variables describing this phenomenon. Both a classical version of the measure and a modified version incorporating spatial dependencies are presented. The results indicate that the use of taxonomic measures allows for a more comprehensive assessment of crime risk than individual indicators and reduces the influence of one-off fluctuations in the data on analytical outcomes.

• Details
• Contents