DISCUSSION
Findings
Conclusions
o By observing kernel map, a crime hotspot was found in the East side of city core.
o Overall, percent poverty, light density, percent aged 10-19, number of parks were the combined best variables to explain the distribution of crime.
o Overall, canopy cover was negatively correlated with crime rate
o In different OLS models, light density and number of parks are consistently strong predictors of high crime.
o Canopy GWR: can best explain crime in residential areas. Low r squared values in commercial and industrial / mixed use regions
o Light GWR: can best explain crime in residential areas in south of Portland
o Parks GWR : can best explain crime in commercial areas, and mixed use areas
o Outdoor lighting and facilities (parks) were not found to deter crime significantly.
o The Canopy GWR can best explain crime in residential areas: canopy is variable and more residential vegetation is usually associated with higher income neighbourhoods.
o Further research could examine the socioeconomic context of the specific region of downtown with a high crime risk.