The development of local calibration factors for implementing the highway safety manual in Maryland.

PDFTech ReportMD-14-SP209B4JSP209B4JHighway safetyTraffic accidentsFatalitiesAccident dataComputer modelsMathematical predictionCalibrationData collectionSamplingMarylandMaryland. State Highway AdministrationShin, HyeonshicLee, Young-JaeDadvar, SeyedehsanMorgan State UniversityUnited States. Dept. of Transportation. Research and Innovative Technology AdministrationUS Transportation CollectionThe goal of the study was to determine local calibration factors (LCFs) to adjust predicted motorvehicle traffic crashes for the Maryland-specific application of the Highway Safety Manual(HSM). Since HSM predictive models were developed using data from states other thanMaryland, the predictive models should be calibrated. LCFs for all 18 facility types werecalculated using Maryland data. After data collection and compilation, samples were drawn basedon 90% confidence level, which increased the overall sample size. After the comparison betweenHSM default crash proportion and Maryland specific data, the use of the Maryland data wassuggested. In general, LCFs for all facilities were less than 1.0, implying Maryland had fewercrashes than predicted crashes estimated by HSM crash prediction methodology. Especially,LCFs for intersections were extremely low. Due to potential issues with unreported minor andproperty damage only crashes, the authors recommend using LCFs for fatal and injuries crasheswhere available. The report concludes with a discussion on data limitation, thoughts about HSMsampling requirement, and suggestions for future research. It should be noted that this study onlyconsidered roadways maintained by SHA and Baltimore City was not part of this study