Time-of-day variations and temporal instability of factors affecting injury severities in large-truck crashes

Using the data from large-truck crashes in Los Angeles over an eight-year period (January 1, 2010 to December 31, 2017), the variation in the influence of factors affecting injury severities during different time periods of the day (morning and afternoon) and from year to year is studied. To capture potential unobserved heterogeneity, random parameters logit models with heterogeneity in the means and variances of the random parameters were estimated considering three possible crash injury-severity outcomes (no injury, minor injury, and severe injury). Likelihood ratio tests were conducted to assess the transferability of model estimation results from different times of the day and from year to year. Marginal effects of the explanatory variables were also calculated to investigate the stability of individual parameter estimates on injury-severity probabilities across time-of-day/time-period combinations. A wide range of parameters were considered including drivers' characteristics, driver actions, truck's characteristics, weather and environmental conditions, and roadway attributes. The results show instability in the effect of factors that influence injury severities in large-truck vehicle crashes across daily time periods and from year to year. However, there were several variables that exhibited relatively stable effects on injury-severity probabilities including driver ethnicity, crashes occurring while backing, sideswipe crashes, hit-object crashes, parked-vehicle crashes, fixed-object crashes, and truck-driver at fault crashes. The findings of this study should be useful for decision makers and trucking companies to better regulate truck operations by time of day. (C) 2019 Elsevier Ltd. All rights reserved.