摘要
为阐明道路交通环境在车祸的预防和控制中的作用,采用典则相关分析法探讨车辆、道路和运输等状况与车祸之间的关系。结果显示,第一典则相关系数为0.9486,主要反映了城市机动化程度与车祸发生数、死亡率和受伤率之间的正相关性,即机动化程度越高,车祸发生数、死亡率和受伤率也越大;第二典则相关系数为0.9220,主要反映了客运量、货运量与车祸发生数之间的正相关性,即客运量和货运量越大,交通事故的发生次数越多;第三典则相关系数为0.6446,主要反映了道路质量与万车事故率之间的相关关系,道路质量好,则万车事故率低。表明道路交通状况与车祸之间存在密切的关系。通过发展道路的基础建设,改善交通状况,可以降低车祸的发生率和伤亡率。典则相关可以分析两组变量间的相关程度,并可进一步分析典则变量与原变量的亲疏关系,这种方法所提供的信息量大,适用范围广,应用计算饥运算方法简便,可用于流行病学研究中对两组随机变量的相关性分析。
In order to expound the impact of traffic condition to road injury, the relations of road injury with motorization, road situation e and highway capacity were studied, using Canonical Correlation analysis method. Results showed that the Canonical Correlation Coefficient A(0. 9486) was representing the direct eorrelation of the motorization with the injury and death rates, referring the higher the motorization, the greater the death and injury rates. The Canonical Correlation Coefficient B(0. 9220) indicated the direct correlation between the highway capacity and the frequency of road injury, as well as the larger the highway capacity, the more frequency of road injury occurrance. The Canonical Correlation Coefficient C(0.6446) revealed the relations between levels of road situation and the accident rates per 10 000 vehicles. Results showed that the higher the quality of roadway, the lower was the accident rate. In view of this, traffic condition was closely related to road injury. Thus, the frequency of accident, death rate and injury rate could be reduced through improving the road quality and traffic conditions. Relations between two sets of random varibales, the ties of the canonical variables and the original variables could all be well analysed with Canonical Correlation analysis. Canonical Correlation is useful to analyse two sets of random variables in epidemiological studies since it provides a great amount of information to be applied and operated widely and easily.
出处
《中华流行病学杂志》
CAS
CSCD
北大核心
1998年第4期227-230,共4页
Chinese Journal of Epidemiology
基金
国务院侨务办公室重点科研基金
广东省卫生厅科研基金
关键词
车祸
交通状况
典则相关分析
Road injury Traffic condition Canonical correlation
