Home
Resource
The impact of indeterminate methods and results on decision-making in highway safety: spatial factors, model specification and measurement error
PDF
PDF format is widely accepted and good for printing.
Plug-in required
PDF-1(1.57 MB)
Citation & Export
View Usage Statistics
Staff View

Citation & Export
Hide

Simple citation

Adediji, Olayemi. The impact of indeterminate methods and results on decision-making in highway safety: spatial factors, model specification and measurement error. Retrieved from https://doi.org/doi:10.7282/t3-7pqq-c570

Export

Click here for information about Citation Management Tools at Rutgers.

Statistics
Hide

Description

TitleThe impact of indeterminate methods and results on decision-making in highway safety: spatial factors, model specification and measurement error
NameAdediji, Olayemi (author); Noland, Robert (chair); Rutgers University; School of Graduate Studies
Date Created2019
Other Date2019-05 (degree)
SubjectCrash frequency, Planning and Public Policy, Traffic accidents -- Mathematical models
Extent1 online resource (x, 207 pages) : illustrations
DescriptionIn highway safety practice, the conventional approach to improving safety by implementing road safety treatments is to use deterministic crash modification factors (CMF). The problem with these crash modification factors is that they are actually indeterminate. They are based on estimates of the associations of various road geometry attributes with crash frequency, which are derived from models that can be affected by many methodological and data problems. And since the methods employed are largely dependent on the analyst’s discretion, results can vary quite widely. The areas of model specification and data are two areas that are very much subject to the analyst’s discretion. Specifying crash models wrongly (introducing specification error) and using data with availability and quality problems (introducing measurement error) can cause erroneous inferences to be made about the results of safety countermeasures that are applied to highway segments, compounding the problem of crash occurrences and potentially creating inefficiency in road safety spending. This problem is exacerbated by the fact that they may be propagated as an industry standard through the existence of an expert manual. I have examined the specific specification error problem of omitted variable bias, where the associations of variables included in safety models are biased due to the omission of certain important variables, and the measurement error problems of data availability and quality. In examining these problems, I used a comparison method, where I compared the results of statistical models affected by the specification and measurement error problems to models where I have attempted to rectify the problems in order to see if there is an improvement in the results of the latter. My findings are mixed; results show no substantial change in the associations with crash frequency between models affected by specification and measurement error and models unaffected for certain variables and show notable change in association for other variables. I have also examined the implications of these problems in practice through interviews of safety practitioners and found that practical limitations preventing transportation agencies from addressing these problems exist.
NotePh.D.
NoteIncludes bibliographical references
Genretheses, ETD doctoral
Persistent URLhttps://doi.org/doi:10.7282/t3-7pqq-c570
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Version 8.5.5
Rutgers University Libraries - Copyright ©2024