Role of Attitudes and Identification in Safety Research

An Empirical Assessment of the Role of Attitudes and Identification in Safety Research

Principal Investigator: Fred L. Mannering, Professor, Department of Civil and Environmental Engineering
Co-Principal Investigator: Michael Maness, Assistant Professor, Department of Civil and Environmental Engineering
Project Duration: 12 months
​Project Budget (Federal UTC Funds): $74,382
Project Budget (Cost-share): $37,192
Institution: University of South Florida

Abstract
Research in highway safety has struggled to deal adequately with two issues: the role that attitudes may play on risk perception and resulting crash and injury-severity likelihoods; and the issue of identification in safety modeling caused by self-selective sampling of safety data (the fact that riskier drivers are likely to be over-represented in crash data bases). This study addresses these two points by first collecting data that focused on highway safety perceptions and using observed crash data from the state of Florida. Statistical model estimation results that address the question of how vehicle usage has changed post-pandemic indicate that safer drivers have reduced their vehicle usage significantly more than riskier drivers. With a greater proportion of vehicle miles travelled now being riskier drivers, this has likely resulted in fundamental shift in injury severity probabilities in observed crashes, which is confirmed using Florida crash data from 2019 (pre-pandemic) and 2020 (pandemic) where a statistically significant shift was found in the factors that determined injury severities in highway the two years. The results indicate that the role of safety attitudes and perceptions are important considerations in the analysis of highway crash data and must be considered in highway-safety practice. The results also indicate the need to focus attention on identification in highway-safety practice because observed statistical estimates may not be entirely due to the effect of specific explanatory variables but may be in part due the change in the mix of risky and safe drivers, a possibility that has important policy implications.

Research Products and Implementation

Scope of Work

​Final Report