Figure - available from: Journal of Geographical Systems
This content is subject to copyright. Terms and conditions apply.
![Posterior medians of the higher-level neighbourhood (exp(γ1)) (a), electoral ward expλωj2+γ2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left( {\exp \left( {\lambda \omega_{{j_{2} }} + \gamma_{2} } \right)} \right)$$\end{document} (b), and police patrol zone (exp(γ3)) terms (c)](publication/334903457/figure/fig3/AS:960346093273129@1605975759973/Posterior-medians-of-the-higher-level-neighbourhood-expg1a-electoral-ward.png)
Posterior medians of the higher-level neighbourhood (exp(γ1)) (a), electoral ward expλωj2+γ2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left( {\exp \left( {\lambda \omega_{{j_{2} }} + \gamma_{2} } \right)} \right)$$\end{document} (b), and police patrol zone (exp(γ3)) terms (c)
Source publication
Characteristics of the urban environment influence where and when crime events occur; however, past studies often analyse cross-sectional data for one spatial scale and do not account for the processes and place-based policies that influence crime across multiple scales. This research applies a Bayesian cross-classified multilevel modelling approac...
Citations
... Bayesian cross-classified multilevel approach to spatiotemporal analysis can help identify geographical contexts that are more salient for local government and public health institutions to establish overdose prevention intervention. Cross-classified multilevel models (Quick, 2019) are useful for measuring the Bayesian spatiotemporal patterns of polysubstance use involved overdoses, allowing for the integration of multiple overlapping geographic contexts, estimate the simultaneous impact of both observed and latent risk factors at multiple spatial levels, and determine the extent to which geographic units of interest account for the variation in overdose over time (Quick, 2019). Furthermore, this approach addresses the Modifiable Areal Unit Problem (MAUP) and the related Modifiable Temporal Unit Problem (MTUP) by leveraging the hierarchical structure to distinguish between variations within and between levels (Jones et al., 2018), incorporating partial pooling and shrinkage in areas where there is sparsity in the data to borrow strength from larger areas or time periods, and allows for the flexible inclusion of random effects for both spatial and temporal scales (Jones et al., 2018). ...
... Bayesian cross-classified multilevel approach to spatiotemporal analysis can help identify geographical contexts that are more salient for local government and public health institutions to establish overdose prevention intervention. Cross-classified multilevel models (Quick, 2019) are useful for measuring the Bayesian spatiotemporal patterns of polysubstance use involved overdoses, allowing for the integration of multiple overlapping geographic contexts, estimate the simultaneous impact of both observed and latent risk factors at multiple spatial levels, and determine the extent to which geographic units of interest account for the variation in overdose over time (Quick, 2019). Furthermore, this approach addresses the Modifiable Areal Unit Problem (MAUP) and the related Modifiable Temporal Unit Problem (MTUP) by leveraging the hierarchical structure to distinguish between variations within and between levels (Jones et al., 2018), incorporating partial pooling and shrinkage in areas where there is sparsity in the data to borrow strength from larger areas or time periods, and allows for the flexible inclusion of random effects for both spatial and temporal scales (Jones et al., 2018). ...
In the current wave of the opioid epidemic, the prevalence of polysubstance use continues to complicate drug-related deaths. Most studies to date use non-spatial statistical approaches to examine the association between polysubstance use and overdose risk, without considering the spatial distribution of these latent sub-patterns of use. This paper describes the utility and potential impact of using disease mapping and Bayesian spatiotemporal approaches for analyzing and monitoring polysubstance use and overdose risk to better respond to the ongoing opioid epidemic. We discuss the application of Bayesian spatiotemporal approaches in analyzing polysubstance use among people who use drugs. Bayesian spatiotemporal analyses offer a salient approach to detecting localized distributions of overdose events and tailor local interventions to community needs in order to reduce polysubstance use and related adverse health among people who use drugs. This can help improve precision and efficacy response in reducing polysubstance use adverse outcomes and optimize resource allocation.
... The spatial frequentist techniques include the zero-inflated negative binomial model (Liu et al., 2018;Swartout et al., 2015), geographically weighted negative binomial regression (GWNBR) (Chen et al., 2020;Wang et al., 2017), spatial Durbin (R. P. Haining & Li, 2020), and spatial spline regression models (Sangalli et al., 2013). Additionally, with the advancement of computational power, Bayesian spatial techniques have gained considerable popularity, such as the Bayesian Poisson hierarchical regression (Law & Haining, 2004;Law & Quick, 2013;Persad, 2020;Quick et al., 2017), Bayesian semiparametric joint quantile regression (Bresson et al., 2021;Chen & Tokdar, 2021;Jang & Wang, 2015;Kottas & Krnjajić, 2009), Bayesian cross-classified multilevel spatial (and temporal) modeling (Quick, 2019) and Bayesian spatial network learning (Baumgartner et al., 2005;Mahmud et al., 2016). Each of these techniques is applied considering different aspects of crime and poses its own advantages and disadvantages. ...
Background setting
Traditional spatial or non-spatial regression techniques require individual variables to be defined as dependent and independent variables, often assuming a unidirectional and (global) linear relationship between the variables under study. This research studies the Bayesian shared component spatial (BSCS) modeling as an alternative approach to identifying local associations between two or more variables and their spatial patterns.
Methods
The variables to be studied, young offenders (YO) and violent crimes (VC), are treated as (multiple) outcomes in the BSCS model. Separate non-BSCS models that treat YO as the outcome variable and VC as the independent variable have also been developed. Results are compared in terms of model fit, risk estimates, and identification of hotspot areas.
Results
Compared to the traditional non-BSCS models, the BSCS models fitted the data better and identified a strong spatial association between YO and VC. Using the BSCS technique allowed both the YO and VC to be modeled as outcome variables, assuming common data-generating processes that are influenced by a set of socioeconomic covariates. The BSCS technique offered smooth and easy mapping of the identified association, with the maps displaying the common (shared) and separate (individual) hotspots of YO and VC.
Conclusions
The proposed method can transform existing association analyses from methods requiring inputs as dependent and independent variables to outcome variables only and shift the reliance on regression coefficients to probability risk maps for characterizing (local) associations between the outcomes.
... Since we aim to compare fatality across countries in the territory under study, a population variable was created and used as offset since the occurrence of these events has a strong positive relationship with the population size (Quick, 2019). The population for each country was obtained from the World Bank database, openly available at https://data.worldbank.org. ...
... Patrones espacio-temporales de delincuencia multiescala: un enfoque bayesiano para modelos multinivel de clasificación cruzada (Quick, 2019) El objetivo es demostrar los datos que se pueden obtener de los procesos multiescala ya que estas toman en cuenta otras escalas espaciales que están relacionadas al crimen como por ejemplo características sociodemográficas, políticas y del entorno. Gracias a una examinación de los patrones delictivos a través del modelo multinivel se puede obtener una mejor comprensión teórica de los procesos multiescala que tienen relación con el delito y pueden orientar a una política de prevención. ...
En la sociedad de hoy los delitos vienen incrementándose y particularmente en la ciudad de Bogotá, lo que ha causado muchos inconvenientes a la Policía Nacional de Colombia, así como también a los centros de seguridad ciudadana. Ante esta situación, se ha propuesto una predicción de tiempo-espacio en los puntos críticos de crímenes y delitos, con la ayuda de inteligencia artificial. Por consiguiente, este trabajo tiene como objetivo analizar, resumir, interpretar y evaluar las distintas técnicas de predicción espacio-temporal de la delincuencia con un panorama inteligente. Por la propia naturaleza de la investigación, se utilizó una metodología de enfoque descriptivo-cualitativo, con la cual se diseñaron fichas de observación estructurada para sistematizar información de cinco bases de datos: Scopus, Web of Science, IEEE, ACM, Springer; dichas publicaciones comprenden desde 2019 hasta junio de 2021. En consecuencia, se encontraron en total 3015 estudios, después del proceso de cribado y verificación de los criterios de exclusión e inclusión, se seleccionaron 132 artículos, luego se aplicaron preguntas Psicólogo Interno Residente (PIR), quedando así 18 artículos. Los principales hallazgos encontrados indican que los algoritmos de redes neuronales resultaron ser uno de los métodos más eficaces para la detección de puntos críticos de delincuencia, dado que los grandes avances de la tecnología coadyuvarían en los próximos años a predecir de forma rápida y eficaz los actos delictivos y los crímenes ubicados en cualquier región del continente latinoamericano.
... Ref. [27] discussed the multi-scale representation of battlefield situation. Ref. [28] proposed a multi-level model to explore the spatial-temporal patterns of crime in different spatial scales of area. They provide guidance for the construction of cognitive models of ship behavior. ...
Ship behavior is the semantic expression of corresponding trajectory in spatial-temporal space. The intelligent identification of ship behavior is critical for safety supervision in the waterborne transport. In particular, the complicated behavior reflects the long-term intentions of a ship, but it is challenging to recognize it automatically for computers without a proper understanding. For this purpose, this study provides a method to model the behavior for computers from the perspective of knowledge modeling that is explainable. Based on our previous work, a semantic model for ship behavior representation is given considering the multi-scale features of ship behavior in cognitive space. Firstly, the multi-scale features of ship behavior are analyzed in spatial-temporal dimension and semantic dimension individually. Then, a method for multi-scale behaviors modeling from the perspective of semantics is determined, which divides the behavior scale into four sub-scales in cognitive space, considering spatial and temporal dimensions: action, activity, process, and event. Furthermore, an ontology model is introduced to construct the multi-scale semantic model for ship behavior, where behaviors with different semantic scales are expressed using the functions of ontology from a microscopic perspective to a macroscopic perspective consecutively. To validate the model, a case study is conducted in which ship behavior with different scales occurred in port water areas. Typical behaviors, which include leveraging the axioms expression and semantic web rule language (SWRL) of the ontology, are then deduced using a reasoner, such as Pellet. The results show that the model is reasonable and feasible to represent multi-scale ship behavior in various scenarios and provides the potential to construct a smart supervision network for maritime authorities.
... Recent studies have indicated micro-places provide a larger contribution to the total spatial variability of crime patterns compared to neighborhoods within cities. These analyses use a hierarchical framework which examines crime across at least two nested spatial units to understand how crime varies between spatial scales (see Hipp, Kim, & Wo, 2021;O'Brien, 2019;O'Brien & Winship, 2017;Quick, 2019;Schnell, Braga, & Piza, 2017). These analyses provide concise estimates of variability which represents a quantification of the unique contribution of a spatial unit of analysis in telling the story of where crimes occur between places. ...
... Our results highlight the role of higher-order spatial units such as neighborhoods and cities when estimating proportions of robbery variability. Like other research that has explored spatial variability, our work further supports findings of nested crime variability with microunits accounting for most of the total variability and with neighborhoods accounting for a small but essential share of the variability (Quick, 2019;Schnell et al., 2017;Steenbeek & Weisburd, 2016). Our results are the first to systematically examine between-city differences of the spatial variability of robbery. ...
Purpose
The law of crime concentration suggests cities have almost no impact on the spatial distribution of crime. This study reinvestigates the relationship between cities and the distribution of crime across the various micro-places and neighborhoods which compose these locations.
Methods
We observed robbery incidents reported to police departments across eight U.S. cities from 2015 to 2019. We calculated the spatial variability of the distribution of robberies attributed to census blocks (i.e., micro-places), census tracts (i.e., neighborhoods), and cities (i.e., macro-places) using variance partitioning techniques with multi-level negative binomial regression models.
Results
Our findings are mixed on the relationship between cities and the spatial distribution of robbery incidents. The descriptive analyses suggest a moderate influence of cities on measures of crime concentration. One of our modeling strategies estimates a larger impact while another strategy observes almost no contribution of cities to the total spatial variability.
Conclusions
This study supports previous research which demonstrates there are overwhelming similarities in the distribution of crime between cities and micro-places account for most of the spatial variability of patterns. While cities do not appear to have a major influence on distributions, future research should continue to clarify these mixed findings and provide a more compelling theoretical account of this relationship.
... After its first use in criminology, GP was applied to biological problems such as the targeting of an infectious disease (Papini and Santosuosso 2016), the prediction of nest locations of bumble bees (Suzuki-Ohno et al. 2010), animal foraging (Le Comber et al. 2006;Raine et al. 2009) and shark hunting patterns (Martin et al. 2009). The obtained results can be compared to those of other analytical methods of mapping the higher or lower probability of crimes occurrence in a given area, as in Quick (2019). ...
Geographic Profiling (GP) attempts to reconstruct the spreading centre of a series of events due to the same cause. The result of the analysis provides an approximated localization of the spreading centre within an area (often represented as a red red), where the probability of finding it is higher than a given threshold (typically 95%). The analysis has as an assumption that the events will be likely to occur at very low probability around the spreading centre, in a ring-shaped zone called the buffer zone. Obvious examples are series of crimes perpetrated by an offender (unwilling to perpetrate offences close to home), or the localities of spread of an invasive species, where the buffer zone, if present, depends on the biological features of the species. Our first aim was to show how the addition of new events may change the preliminary approximate localization of the spreading centre. The analyses of the simulated data showed that if B , the parameter used to represent the radius of the buffer zone, varies within a range of 10% from the real value, after a low number of events (7–8), the method yields converging results in terms of distance between the barycentre of the red zone and the “real” user provided spreading centre of a simulated data set. The convergence occurs more slowly with the increase in inaccuracy of B . These results provide further validity to the method of the GP, showing that even an approximate choice of the B value can be sufficient for an accurate location of the spreading centre. The results allow also to quantify how many samples are needed in relation to the uncertainty of the chosen parameters, to obtain feasible results.
... The Bayesian spatiotemporal model has been widely applied in epidemiological studies and disease mapping [48]. Crime studies employing this model have also emerged in Western literature [21][22][23]47,[49][50][51][52]. However, very few crime studies applying the Bayesian approach have been conducted in a Chinese context, except for several case studies in Wuhan, China [25,53,54]. ...
Chinese cities have been undergoing extraordinary changes in many respects during the process of urbanization, which has caused crime patterns to evolve accordingly. This research applies a Bayesian spatiotemporal model to explore and understand the spatiotemporal patterns of crime risk from 2008 to 2017 in Changchun, China. The overall temporal trend of crime risk, the effects of land use covariates, spatial random effects, and area-specific differential trends are estimated through a Bayesian spatiotemporal model fitted using the Integrated Nested Laplace Approximation (INLA). The analytical results show that the regression coefficient for the overall temporal trend of crime risk changed from significantly positive to negative after the land use variables are incorporated into the Bayesian spatiotemporal model. The covariates of road density, commercial and recreational land per capita, residential land per capita, and industrial land per capita are found to be significantly associated with crime risk, which relates to classic theories in environmental criminology. In addition, some areas still exhibit significantly increasing crime risks compared with the general trend even after controlling for the land use covariates and the spatial random effects, which may provide insights for law enforcement and researchers regarding where more attention is required since there may be some unmeasured factors causing higher crime trend in these areas.
... However, in some instances it may make sense to consider an additive measurement error structure, represented as, * = + , where ( ) ≠ 0 if the errors are systematic. This is the case when crime rates are log-transformed; a common strategy used to normalise their often right-skewed distribution (Sutherland et al., 2013;Whitworth, 2011), to interpret effects in relative terms (Goulas & Zervoyianni, 2013;Witt & Witte, 2000), or as a result of employing generalised linear models where logs are used as the link function, such as Poisson models (Quick, 2019;Sampson et al., 1997). Crucially, log-transforming crime rates also has the effect of transforming the observed multiplicative measurement error into an additive mechanism, since: log( * ) = log( ) = log( ) + log( ). ...
Objectives: Assess the extent to which measurement error in police recorded crime rates impact the estimates of regression models exploring the causes and consequences of crime.
Methods: We focus on linear models where crime rates are included either as the response or as an explanatory variable, in their original scale, or log-transformed. Two measurement error mechanisms are considered, systematic errors in the form of under-recorded crime, and random errors in the form of recording inconsistencies across areas. The extent to which such measurement error mechanisms impact model parameters is demonstrated algebraically, using formal notation, and graphically, using simulations.
Results: Most coefficients and measures of uncertainty from models where crime rates are included in their original scale are severely biased. However, in many cases, this problem could be minimised, or altogether eliminated by log-transforming crime rates. This transforms the multiplicative measurement error observed in police recorded crime rates into a less harmful additive mechanism.
Conclusions: The validity of findings from regression models where police recorded crime rates are used in their original scale is put into question. In interpreting the large evidence base exploring the effects and consequences of crime using police statistics we urge researchers to consider the biasing effects shown here. Equally, we urge researchers to log-transform crime rates before they are introduced in statistical models.
... A recent study used Bayesian cross-classified multilevel modeling to account for the processes and place-based policies that influence violent crime across multiple scales. The study found that violent crime is positively associated with population size, residential instability, the central business district, and commercial, government-institutional, and recreational land uses within small areas, and negatively associated with civic engagement within higher unit, electoral wards (Quick, 2019). ...
Scholars argue that housing abandonment increases area criminal activity. The link between abandoned properties and crime has led to the assumption that demolition of abandoned properties will stymie critical activity and thus improve neighborhood safety. Although cities spend millions of federal and local funds on demolitions every year, very little research has explored the empirical effects of demolitions on crime. Does demolition lead to a reduction in nearby crime? This study answers this question by quantifying the relationship between abandoned building demolition programs and nearby crime using a difference-in-difference approach on 559 abandoned buildings demolished in Kansas City, Missouri, between 2012 and 2016. This study finds that demolition of abandoned properties does not have any significant impact on nearby violent and property crime. This analysis shows that a change in nearby crime is attributable to differences in nearby socioeconomic and housing characteristics, rather than to the demolition of abandoned properties.
