PROCEEDINGS OF THE INSTITUTION OF CIVIL ENGINEERS-TRANSPORT, cilt.168, sa.6, ss.532-542, 2015 (SCI-Expanded)
This study proposes a spatial analysis for road traffic accidents (RTAs) using geographical information system technology based on the number of RTAs and the length of the roads in each province of Turkey. Spatial autocorrelation analysis was performed to show whether the provinces with high mortality rates show clustering. Using spatial autocorrelation analysis the spatial distribution of provinces with high mortality rates was found to be non-random and to be clustered with a significance of p < 0.05. Factor analysis was carried out using population, population density, number of vehicles, number of vehicle-kilometres, number of passenger-kilometres and number of ton-kilometres. The results show that mortality rates based on the number of RTAs are higher in undeveloped provinces compared with developed provinces, and that mortality rates based on the length of roads are higher in developed provinces than in undeveloped provinces.