Statistical Analysis Of Canadian Traffic Accidents

As the most common mode of transportation in the world,car plays a significant role our daily live.According to statistics,in the world,one person dies from a car accident every 25 seconds.How to reduce the accident has become a topic been paid highly attention to.This paper aims to find out the factors leading to road accidents and through statistical analysis.Due to the random characteristics of road traffic accidents,using multivariate statistical methods,it is possible to analyze the significant risk factors on accidents directly and effectively,and create effective improvement measures.Based on the Haddon matrix developed by William Haddon in 1970,this paper analyzes accidents by human,vehicle,road and environmental factors.A number of accident statistical indicators are used to describe the distribution of 3400,852 traffic accidents in Canada from 2000 to 2009 and the negative binomial regression model is used to analyze the significant correlation among 64 factors.Among human factors,drunk driving is the most significant factor causing car accidents;the possibility of accidents caused by speeding is the highest;the probability of accidents caused by lighting failure factors is the highest in vehicle failure factors;the number of accidents on dry roads are higher: among the road shape factors,flat straight roads are the most prone to accidents;no traffic signals in intersections is the most important cause of accidents;accidents occur more during daylight savings time;snow weather has the greatest impact on accidents.This research can be used to improve the design of road and traffic engineering facilities,provide suggestions for the implementation of traffic safety management by government departments,help design traffic safety education content for motor vehicle drivers.This paper examine traffic accidents data in Canada throughout 2000-2009,from National Collision Database(NCDB).