A Bayesian procedure for evaluating the frequency of calibrationfactor updates in highway safety manual (HSM) applications

abstract

The Highway Safety Manual (HSM) presents statistical models to quantitatively estimate an agency’ssafety performance. The models were developed using data from only a few U.S. states. To account forthe effects of the local attributes and temporal factors on crash occurrence, agencies are required to cal-ibrate the HSM-default models for crash predictions. The manual suggests updating calibration factorsevery two to three years, or preferably on an annual basis. Given that the calibration process involvessubstantial time, effort, and resources, a comprehensive analysis of the required calibration factor updatefrequency is valuable to the agencies. Accordingly, the objective of this study is to evaluate the HSM’srecommendation and determine the required frequency of calibration factor updates. A robust Bayesianestimation procedure is used to assess the variation between calibration factors computed annually, bien-nially, and triennially using data collected from over 2400 miles of segments and over 700 intersectionson urban and suburban facilities in Florida. Bayesian model yields a posterior distribution of the modelparameters that give credible information to infer whether the difference between calibration factorscomputed at specified intervals is credibly different from the null value which represents unaltered cali-bration factors between the comparison years or in other words, zero difference. The concept of the nullvalue is extended to include the range of values that are practically equivalent to zero. Bayesian inferenceshows that calibration factors based on total crash frequency are required to be updated every two yearsin cases where the variations between calibration factors are not greater than 0.01. When the variationsare between 0.01 and 0.05, calibration factors based on total crash frequency could be updated everythree years.