A state transportation agency wants to evaluate engineering countermeasures applied to a two-lane, two-way rural highway segment. This 4-mile segment has been identified as having 14, 19, and 23 total crashes in the last three years, before the implementation of the countermeasures. The AADT at the segment for the last three years is 5,630 vehicles per day, and the AADT for after implementation is 6,240 vehicles per day. Compute the expected average crash frequency for the after period, assuming the treatment was not implemented. The segment has the same conditions as the base conditions established for the SPF.

Question

What will be an ideal response?

Answer 1

Solution: Use the Empirical-Bayes Method for this problem.

(a) Step 1: Compute the predicted average crash frequency for site type x for each year of the before period using an appropriate SPF, which is Eq. 5.17:



The predicted average crash frequency is given by Equation 5.53:



As the segment conditions are the same as the base conditions for the SPF, CMFs and the calibration factor are assumed as 1. Therefore:



And the predicted average crash frequency in the before period will be:



(b) Step 2: Compute and sum up the expected average crash frequency for the segment over the before period.

The expected average crash frequency is obtained from the predicted crash frequency and the weight, the latter of which is based on the overdispersion parameter and is given by Equation 5.54:



The weight factor is given by Equation 5.55:



For rural two-lane, two-way roads, the overdispersion parameter is given by

Equation 5.17:



And the weight factor will be:



The expected average crash for the before period will be:



(c) Step 3: Compute the predicted average crash frequency for the segment for each

year during the after period, assuming that the treatment was not implemented.

The predicted average crash frequency for site type x for each year of the after

period is given by Eq. 5.17



The predicted average crash frequency is given by Equation 5.53:



As the segment conditions are the same as the base conditions for the SPF, CMFs

and the calibration factor are assumed as 1. Therefore:



And the predicted average crash frequency in the after period will be:



(d) Compute an adjustment factor. Use Equation 5.56:



(e) Compute the expected average crash frequency for the segment in the after

period, assuming the treatment was not implemented: