TRANSFERABILITY of CRASH MODIFICATION FACTORS via GRAPHICAL CAUSAL MODELS: AN INTRODUCTION

• Activity Number: 424 - Contributed Poster Presentations: Transportation Statistics Interest Group
• Type: Contributed
• Date/Time: Tuesday, July 30, 2019 : 2:00 PM to 3:50 PM
• Sponsor: Transportation Statistics Interest Group
• Author(s): Gary Davis* and Jingru Gao
• Companies: University of Minnesota and University of Minnesota
• Keywords: Crash Modification Factor; graphical model; causal inference

Abstract

This paper describes exploratory analyses of how to transfer a crash modification factor estimated under one situation to other situations, which is likely to be more important as automated vehicles improve their capabilities and increase their market share. The authors work with graphical models describing the dependencies among a set of relevant variables in a crash mechanism, and focus on (1) identifying sufficient conditions for taking causal information determined in one situation and applying to another, and (2) deriving expressions for computing the transferred quantities. Three simplified but plausible scenarios have been proposed. For each scenario, first, transportability analyses developed by Pearl and his associates are used to develop a re-calibration formula where an existing CMF can be adjusted to reflect new conditions, and second, computation examples are proposed to illustrate those results.