Abstract
In this paper, to study the frequency and severity of traffic accidents, a combination of geographic information system and spatial analysis has been used. For this purpose, methods of planar and network statistics including kernel density estimation and nearest neighbor distance analysis were applied for three types of accidents including fatal, injury, and material property damages. In addition, drawbacks of the traditional planar point pattern and the benefits of network analysis are offered. The network statistics are executed through SANET version 4 and ArcMap on the Semnan–Garmsar road in Iran. The nearest neighbor output analysis is consist of four curves, and regarding the fact that the amounts for accidents resulting in fatalities, injuries, and damages, the observed value curve is above the 5% confidence interval; accidents in the study region are more clustered than would be expected by random chance.