The objective of this study was to determine the change in crash history at intersections after installation of a traffic signal. Signals installed based on the warrants from an engineering study resulted in a decrease in angle collisions with an increase in the less severe rear end collisions. At intersections where signal warrants were not met, there was a smaller decrease in angle collisions and a larger increase in rear end collisions. The removal of unwarranted traffic signals did not result in a crash problem. The analysis shows that the warrants given in the MUTCD should be followed when determining the need for a traffic signal installation with the installation based on an engineering study.
Available at: http://works.bepress.com/eric_green/9/