USE OF STATISTICAL METHODS FOR HIGHWAY SAFETY (CE 434 /CE 534) Purposes: 1. To identify problematic highway sections. 2. To decide whether an observed high/low value of accidents is due to random flactution nature or there is something unusual regarding the road and its environment. 3. To evaluate the significance of any treatment or change applied to any section of a highway. 1) Based on: Comparisons of different sets of statistics. Example (Homework): There are 20 equal length of sections (intersections) within a particular highway network. The following table gives the accident record for 3 years period: Calculate norm; variance (standart deviation); cofficient of variance and decide which highway sections should be investigated in detail? Section Accident Total 1 33 2 21 3 36 4 42 5 8 6 90 7 17 8 22 9 33 10 18 11 22 12 18 13 13 14 18 15 76 16 14 17 21 18 28 19 22 20 35 2) Statistical Tests (Poisson’s Test) Example (Homework): The accident record over the last 8 years are given for a highway section. Whether 7 accidents in year 8 happened by chance (random flactuation nature of accidents) or whether there has been some change at the site which means that the next year level will remain high? Year Accident/ Year 1 3 2 5 3 3 4 6 5 5 6 4 7 7 8 7