Abstract
An improved Safety Index Model is developed for narrow bridges using a statistical approach when bridge and approach roadway characteristics are known. Information was collected regarding the variables such as bridge width, approach roadway width, length, sight distance, 85% approach speed, shoulder reduction, grade continuity, average daily traffic, type of traffic control at an accident site, environmental and road conditions at the time of the accident, number of persons injured, number of persons killed in an accident, alignment and curvature at the stie. Cluster analysis was used to classify the 78 bridges into more-safe and less-safe groups. Correlation and factor analyses were employed to find the interrelationships between the independent variables of bridge and approach roadway characteristics. Stepwise regression was used to find the most important variables related to accident rate. On the groups developed by cluster analysis, logistic regression was employed to develop a model that is a function of several variables which were found to be significantly related to accident rate. The enhanced Safety Index Model consists of the variables: bridge width, length, average daily traffic, and speed as well as the subjective safety factors F(,6) (grade continuity), F(,7) (shoulder reduction) and F(,9) (traffic mix). The model was arrived at scientifically, using more objective procedures of classification and has a stronger correlation with the accident rate than the previous models by others. It is also more parsimonious in variables. The fewer variables yield a much higher coefficient of multiple determination (R('2)) compared to the previous models when accident rate is used as the dependent variable. It is very sensitive to changes or improvements in the constituent factors. The model yields the fraction of concordant pairs of predicted probabilities and responses as 0.91 and a high rank correlation of 0.81 between predicted probability and response which indicates the goodness of the model. It also gives the safety index directly and can be used to identify a potentially hazardous narrow bridge.
Gandhi, Radhakumari Bodapati Venkata (1983). A statistical approach to the development of an improved safety index model for narrow bridges. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -574979.