Date of Award

Fall 1997

Document Type


Degree Name

Master of Science in Transportation - (M.S.)


Executive Committee for the Interdisciplinary Program in Transportation

First Advisor

Kyriacos Mouskos

Second Advisor

I-Jy Steven Chien

Third Advisor

Louis J. Pignataro


This thesis presents an analysis of reported accidents on multilane highways and regression models that identify the primary explanatory variables that have a significant effect on accident rates midblock to signalized intersections. The analysis was based on traffic and reported accident data provided by New Jersey DOT. The access points per mile and the accident rates per million-vehicle-miles-traveled were analyzed based on a large number of roadway sections which were selected from five NJ State routes. Comparative accident analysis related with traffic and roadway geometric characteristics were performed. The analysis showed that approximately 30% of accidents occurred between intersections, which were primarily attributed to the presence of access points. Among these, about 80% of the accidents were caused by a vehicle moving straight through on the mainline and a turning vehicle from/to an access point. Although nonlinear models show good fit, none of the coefficients of the variables show significant t-statistic values. It can be concluded that no good regression models among those tested provide good estimation of accident rates for multilane highways. A field study was conducted and presented several quantitative variables of speed reduction, delay and percentages of affected vehicles at access points.