Abstract
Perceived safety of the built environment—a cognitive assessment different from emotional fear of crime—might affect the number of potential crime victims in an area and thus affect crime opportunities. The perceived safety derived from street view imagery has propelled scholars to examine its relationship with crime. The literature, however, has not addressed the related geographic scale variability issue; that is, the choice of the geographic analytical units might affect the relationship between area-based perceived safety and crime. This study explores how the relationships between street-view-derived perceived safety and both street thefts and street robberies vary by different spatial scales in Cincinnati. Results of negative binomial models show that perceived safety is positively associated with street thefts and street robberies at both the street segment and census block levels, but is negatively associated with these crimes at the census block group level. The relationship is not statistically significant at the census tract level. This variability is explained by the different freedom of avoidance behaviors in response to perceived safety, which change by geographic scale. The research further evaluates the within variance and between variance of perceived safety at different scales. Compared to between variance, within variance is smaller at both the street segment and block levels, but larger at both the block group and tract levels. This variability can be a source of model instability across multiple geographical scales. In short, the multiscale assessment shows that larger spatial units like the census tract are unsuitable for perceived safety–crime analysis.
不同于对犯罪的情感恐惧, 对建成环境安全的感知是一种认知评估, 它可能会影响潜在犯罪受害者数量, 进而影响犯罪机会。从街景影像中获得的感知安全, 促使学者们研究其与犯罪的关系。然而, 文献没有涉及地理尺度的差异性: 面域感知安全与犯罪的关系可能受到地理分析单元的影响。本文以美国辛辛那提市为例, 探讨了从街景获得的感知安全与街头盗窃和抢劫的关系, 如何随空间尺度而变化。负二项模型结果表明, 在路段尺度和人口普查街区尺度, 感知安全与街头盗窃和抢劫呈正相关;在人口普查街区组尺度, 感知安全与街头盗窃和抢劫呈负相关。这种关系在人口普查地区尺度上没有统计学意义。回避感知安全的自由度的不同, 能解释地理尺度的差异性。本文进一步评估了不同尺度下感知安全的尺度内方差和尺度间方差。与尺度间方差相比, 尺度内方差在路段和街区尺度都较小, 在街区组和地区尺度都较大。这种差异性可能是多维地理尺度模型不稳定性的来源。多尺度评估表明, 对感知安全和犯罪关系的分析不适用于人口普查地区等大空间单元。
La seguridad percibida del entorno construido–una ponderación cognitiva diferente del miedo emocional a la criminalidad–podría afectar al número potencial de víctimas de los criminales en una zona y, de esa manera, afectar las oportunidades de delinquir. La seguridad percibida, que se deriva de la imagenería callejera, ha impulsado a los estudiosos a examinar su relación con la criminalidad. Sin embargo, en la literatura no se han abordado cuestiones relacionadas con la variabilidad de la escala geográfica; es decir, la escogencia de las unidades geográficas de análisis podría afectar la relación entre la seguridad percibida basada en zona y la delincuencia. Este estudio explora el modo como las relaciones entre la seguridad percibida, derivada de lo que se ve en las calles, tanto en términos de hurtos como de atracos callejeros, varía en Cincinnati según las diferentes escalas espaciales. Los resultados de los modelos binomiales negativos muestran que la seguridad percibida está positivamente asociada con los hurtos y atracos callejeros, tanto a nivel de segmento de calle como de bloque censal, pero se asocia negativamente con estos delitos a nivel de grupo de bloques censales. La relación no es estadísticamente significativa a nivel de sección censal. Esta variabilidad se explica por la diferente libertad de los comportamientos para evitar la exposición al crimen en respuesta a la seguridad percibida, que cambia según la escala geográfica. También se evalúa en la investigación la varianza interna y la varianza entre las distintas escalas de seguridad percibida. Comparada con la varianza entre áreas, la varianza dentro de éstas es menor tanto a nivel de segmento de calle como de manzana, pero mayor tanto a nivel de grupo de manzanas como de tramo. Esta variabilidad puede ser una fuente de inestabilidad del modelo a través de múltiples escalas geográficas. En resumen, la evaluación multiescalar muestra que las unidades espaciales más grandes, como el tracto censal, no son adecuadas para el análisis de la percepción de seguridad y delincuencia.
Disclosure Statement
The authors declare no conflict of interest
Acknowledgment
We sincerely appreciate anonymous reviewers for their advice to improve the quality of this study. Dr. Lin Liu serves as the corresponding author of the article.
Correction Statement
This article has been corrected with minor changes. These changes do not impact the academic content of the article.
Additional information
Notes on contributors
Hanlin Zhou
HANLIN ZHOU is a PhD Candidate in the Department of Geography and Planning, University of Toronto, Toronto ON M5S 3G3, Canada, and Department of Geography, Geomatics, and Environment, University of Toronto Mississauga, Mississauga ON L5L 1C6, Canada. E-mail: [email protected]. His research interests encompass understanding the environmental impact on human activities, such as health behaviors and crime behaviors.
Lin Liu
LIN LIU (coresponding author) is a Professor in the Department of Geography and GIS at the University of Cincinnati, Cincinnati, OH 45221. E-mail: [email protected]. His research interests include crime mapping and analysis, as well as GIScience and its applications.
Jue Wang
JUE WANG is an Assistant Professor in the Department of Geography, Geomatics, and Environment, University of Toronto Mississauga, ON L5L1C6, Canada. E-mail: [email protected]. His research interests encompass investigating the impacts of the urban environment on human behavior and public well-being, along with the domains of human mobility, GIS, geospatial big data, and spatiotemporal analysis.
Kathi Wilson
KATHI WILSON is a Professor in the Department of Geography, Geomatics, and Environment, University of Toronto Mississauga, Mississauga ON L5L 1C6, Canada. E-mail: [email protected]. Her research focuses on determinants of health and access to health and social services among immigrant populations.
Minxuan Lan
MINXUAN LAN is an Assistant Professor in the Department of Geography and Planning at The University of Toledo, Toledo, OH 43606. E-mail: [email protected]. His research interests include spatial-temporal analysis and its applications to crime, public health, and human sustainability.
Xin Gu
XIN GU is a PhD Candidate in the Department of Geography and GIS at the University of Cincinnati, Cincinnati, OH 45221. E-mail: [email protected]. His research interests include human mobility, big data analysis, crime pattern analysis, and crime prediction.