Application of spatial regression to investigate current patterns of crime in the north of Portugal

Research output: Contribution to conferencePoster

2 Downloads (Pure)

Abstract

Crime is one of the main problems of societies. However, research on this phenomenon has not yet reached a consensus on what factors influence the variability of crime rates and how it positions itself and acts in space. This work aims to contribute to a greater understanding of crime in the north of Portugal by investigating demographic and socioeconomic factors that may be associated with it. We explore the use of spatial statistics techniques through a detailed exploratory spatial data analysis (ESDA) first, and the application of regression models (Ordinary Least Squares and GWR – Geographically Weighted Regression) afterwards. The results show a crime hotspot on the coastline, and spatial homogeneity of low values between central-south municipalities. Crime patterns might be explained by population density, distance to the district capital and beneficiaries of Social Integration Income, who are individuals without socioeconomic conditions that need financial assistance from the Portuguese Government. However, predictions based on a GWR model with these factors may not be appropriate, given the limitations of this technique.
Original languageEnglish
Publication statusPublished - 2017
EventAGILE 2017: 20th conferemce on geo-information science - Wageningen University & Research, Wageningen, Netherlands
Duration: 9 May 201712 May 2017

Conference

ConferenceAGILE 2017
CountryNetherlands
CityWageningen
Period9/05/1712/05/17

Fingerprint

Portugal
offense
regression
financial assistance
crime rate
socioeconomic factors
social integration
demographic factors
population density
municipality
data analysis
statistics
district
income
society

Keywords

  • Crime pattern
  • Spatial non-stationarity
  • Spatial regression
  • ESDA
  • OLS
  • GWR

Cite this

@conference{3873fc935a7f4f36bbf06c4999d973e2,
title = "Application of spatial regression to investigate current patterns of crime in the north of Portugal",
abstract = "Crime is one of the main problems of societies. However, research on this phenomenon has not yet reached a consensus on what factors influence the variability of crime rates and how it positions itself and acts in space. This work aims to contribute to a greater understanding of crime in the north of Portugal by investigating demographic and socioeconomic factors that may be associated with it. We explore the use of spatial statistics techniques through a detailed exploratory spatial data analysis (ESDA) first, and the application of regression models (Ordinary Least Squares and GWR – Geographically Weighted Regression) afterwards. The results show a crime hotspot on the coastline, and spatial homogeneity of low values between central-south municipalities. Crime patterns might be explained by population density, distance to the district capital and beneficiaries of Social Integration Income, who are individuals without socioeconomic conditions that need financial assistance from the Portuguese Government. However, predictions based on a GWR model with these factors may not be appropriate, given the limitations of this technique.",
keywords = "Crime pattern, Spatial non-stationarity, Spatial regression, ESDA, OLS, GWR",
author = "Costa, {Jo{\~a}o Diogo} and Costa, {Ana Cristina}",
note = "Costa, J. D., & Costa, A. C. (2017). Application of spatial regression to investigate current patterns of crime in the north of Portugal. Poster session presented at AGILE 2017, Wageningen, Netherlands.; AGILE 2017 : 20th conferemce on geo-information science ; Conference date: 09-05-2017 Through 12-05-2017",
year = "2017",
language = "English",

}

Application of spatial regression to investigate current patterns of crime in the north of Portugal. / Costa, João Diogo; Costa, Ana Cristina.

2017. Poster session presented at AGILE 2017, Wageningen, Netherlands.

Research output: Contribution to conferencePoster

TY - CONF

T1 - Application of spatial regression to investigate current patterns of crime in the north of Portugal

AU - Costa, João Diogo

AU - Costa, Ana Cristina

N1 - Costa, J. D., & Costa, A. C. (2017). Application of spatial regression to investigate current patterns of crime in the north of Portugal. Poster session presented at AGILE 2017, Wageningen, Netherlands.

PY - 2017

Y1 - 2017

N2 - Crime is one of the main problems of societies. However, research on this phenomenon has not yet reached a consensus on what factors influence the variability of crime rates and how it positions itself and acts in space. This work aims to contribute to a greater understanding of crime in the north of Portugal by investigating demographic and socioeconomic factors that may be associated with it. We explore the use of spatial statistics techniques through a detailed exploratory spatial data analysis (ESDA) first, and the application of regression models (Ordinary Least Squares and GWR – Geographically Weighted Regression) afterwards. The results show a crime hotspot on the coastline, and spatial homogeneity of low values between central-south municipalities. Crime patterns might be explained by population density, distance to the district capital and beneficiaries of Social Integration Income, who are individuals without socioeconomic conditions that need financial assistance from the Portuguese Government. However, predictions based on a GWR model with these factors may not be appropriate, given the limitations of this technique.

AB - Crime is one of the main problems of societies. However, research on this phenomenon has not yet reached a consensus on what factors influence the variability of crime rates and how it positions itself and acts in space. This work aims to contribute to a greater understanding of crime in the north of Portugal by investigating demographic and socioeconomic factors that may be associated with it. We explore the use of spatial statistics techniques through a detailed exploratory spatial data analysis (ESDA) first, and the application of regression models (Ordinary Least Squares and GWR – Geographically Weighted Regression) afterwards. The results show a crime hotspot on the coastline, and spatial homogeneity of low values between central-south municipalities. Crime patterns might be explained by population density, distance to the district capital and beneficiaries of Social Integration Income, who are individuals without socioeconomic conditions that need financial assistance from the Portuguese Government. However, predictions based on a GWR model with these factors may not be appropriate, given the limitations of this technique.

KW - Crime pattern

KW - Spatial non-stationarity

KW - Spatial regression

KW - ESDA

KW - OLS

KW - GWR

M3 - Poster

ER -