Measuring geographies of fear of crime and crime in semi‑rural Sweden

The lack of research on why certain locations are perceived as unsafe has often been attributed to the scarcity of suitable data on place-based fear of crime. To address this gap, this study explores two spatial data collection methods applied in semi-rural areas of Kristianstad, Sweden. The first method uses a digital sketch map integrated with open-ended questions, accessed via an online web map platform, allowing participants to directly mark unsafe locations and resulting in geographically precise data on perceived fear of crime. The second method involves a random sample survey, employing open-ended questions to identify unsafe locations and circumstances, with responses geocoded into points, lines, or polygons for spatial inquiry. The study identifies similarities and differences in the geography of fear of crime derived from these methods. Strong alignment between the web map and survey is observed in both locations, and Spearman’s rank correlation also reveals a significant link between signal crimes and fear of crime. Kernel Density Estimation and Nearest Neighbour analysis reveal significant clusters of fear of crime in both areas, often overlapping with crime hotspots. The findings suggest that a convenience sample collected via a web application performs comparably to a traditional open-ended survey with a random sample, indicating that simpler and more cost-effective methods may be sufficient for capturing place-based fear of crime. This methodological insight supports pragmatic data collection and informs targeted, place-based fear of crime prevention strategies in both rural and urban contexts.