Assessing the Impact of Traffic Volume Changes and Other Confounding Factors in Observational Before–After Safety Studies: a “No Treatment” Evaluation During the COVID-19 Pandemic in Canada
Date
2024-04-18
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
ORCID
Type
Thesis
Degree Level
Masters
Abstract
As motor vehicle accidents result in significant societal and economic costs, it is crucial to evaluate the impact of safety measures accurately. Therefore, this thesis examines the effects of COVID-19-induced traffic volume changes and other factors on current methods used to assess the effectiveness of road safety engineering countermeasures. The study analyzes four observational before-after (BA) study methodologies: comparison group (CG) method, empirical Bayes (EB) method, a combination of EB and CG (EB with CG) method, and full Bayes (FB) method, which are commonly used to isolate the effects of countermeasures from confounding factors such as regression-to-the-mean (RTM), traffic volume fluctuations (exposure), unrelated effects and maturation (time trends). To accomplish this goal, a hypothetical BA study was conducted on multiple untreated signalized intersections across different Canadian jurisdictions during the period 2017-2021, which included COVID-19 pandemic. When these methodologies effectively account for confounding factors, no changes in expected collision frequency should be observed from the pre-pandemic to the post-pandemic period, assuming no specific safety countermeasures were implemented at the study sites.
The findings of this research indicate that among all the BA study types, the crash modification factors (CMFs) consistently exhibited values of approximately 1.0 as treatment effectiveness outcomes (i.e., 0% change in collision frequency from the before to the after period). The results demonstrate that all the examined methods successfully detected the hypothetical treatment within the confidence intervals of the estimated CMFs. However, they exhibited varying levels of accuracy and precision. These findings hold significant importance for practitioners to make informed choices regarding the selection and implementation of assessment methodologies for evaluating countermeasures introduced during COVID-19 pandemic.
Description
Keywords
before-after study, confounding factors, COVID-19, hypothetical treatment, traffic volume changes
Citation
Degree
Master of Science (M.Sc.)
Department
Civil and Geological Engineering
Program
Civil Engineering