@article{ref1,
title="Reliability of crash prediction models: a guide for quantifying and improving the reliability of model results",
journal="NCHRP research report",
year="2021",
author="Srinivasan, Raghavan and Lan, Bo and Mozingo, Caroline and Bonneson, James and Lyon, Craig and Persaud, Bhagwant and Bahar, Geni",
volume="",
number="983",
pages="1-102",
abstract="NCHRP Research Report 983: Reliability of Crash Prediction Models: A Guide for Quan- tifying and Improving the Reliability of Model Results (hereinafter the Guide) describes methods and procedures to (a) quantify the impact of selecting or neglecting certain data parameters in safety estimate predictions, and (b) estimate, interpret, and improve the reli- ability of predictions and use of crash modification factors (CMFs). It also describes several scenarios encountered by practitioners in using crash prediction models (CPMs) and CMFs for safety estimate predictions. The Guide will be of interest to practitioners responsible for making informed decisions about CPMs and CMFs.   The AASHTO Highway Safety Manual (HSM) provides fact-based, analytical tools and techniques to quantify the potential safety impacts of planning, design, operations, and maintenance decisions. Part C of the HSM contains predictive methods for rural two-lane roads, rural multilane highways, and urban and suburban arterials; however, it does not include methods for consistently ensuring model reliability. Since the publication of the HSM, the safety analysis state of practice has progressed and more has been learned about the impact on the accuracy of assumptions made during the development of CPMs using HSM procedures. Practitioners are also striving to fully understand and appropriately com- municate the benefits of the HSM methods and the results derived from these methods.   Under NCHRP Project 17-78, &quot;Understanding and Communicating Reliability of Crash Prediction Models,&quot; the University of North Carolina at Chapel Hill developed the Guide that contains guidelines for (a) quantification of the reliability of CPMs for practitioner use; (b) interpretation of model reliability; and (c) application of CPMs accounting for, but not limited to, assumptions, data ranges, and intended and unintended uses.   The Guide is accompanied by (a) a conduct of research report (published as NCHRP Web-Only Document 303: Understanding and Communicating Reliability of Crash Prediction Models) that details the research activities and methods, (b) a communications plan, (c) a flyer to promote the Guide, and (d) a PowerPoint presentation that summarizes the research effort. These materials are available on the TRB website (www.trb.org) and can be found by searching for &quot;NCHRP Research Report 983.&quot;<p /> <p>Language: en</p>",
language="en",
issn="2572-3766",
doi="10.17226/26439",
url="http://dx.doi.org/10.17226/26439"
}