Are SUVs “Supremely Unsafe Vehicles”?: Analysis of Rollovers and Injuries with Sport Utility Vehicles

Abstract

With increasing speed limits and more light trucks penetrating the market, concern over their rollover risk is growing. In particular, the general public and automobile manufacturers would like to know if the increasingly popular sport utility vehicles (SUVs) are indeed safer than other vehicle platforms. The influences of various vehicle platforms on rollovers and driver injuries were investigated. Specifically, (a) the rollover intensities of SUVs vis-à-vis those of other vehicle types in single-vehicle crashes and (b) the severity of the resulting driver injury were explored. Data from a good-quality federally maintained database were used for crash analysis. The database contains a relatively clean stratified sample of police-reported tow-away crashes nationwide, and it contains detailed information about vehicle rollovers. Rollover intensity, captured by the number of quarter turns, was investigated by using weighted negative binomial models; injury severity, measured on the abbreviated injury scale, was examined by using weighted ordered logit models. New insights emerged about the factors that increase rollover intensity and injury severity. As expected, SUVs are more likely to roll over and therefore injure their occupant drivers more severely. However, SUVs also protect their drivers during collisions because of their greater crashworthiness. In fact, the SUV crashworthiness effect exceeds the rollover effect, on average. The implications of these findings are discussed.

Analysis of driver injury severity in wrong-way driving crashes on controlled-access highways

“…Given the ordered nature of the injury severity in crashes (representing an ordinal outcome), these methodological approaches generally fall under two main categories (based on whether this nature is considered or not): ordered-response models and unordered-response models. Ordered logit/probit (Khattak and Rocha, 2003;Lee and Li, 2014), generalized ordered logit (Wang and Abdel-Aty, 2008;Abegaz et al, 2014), and mixed generalized ordered logit (Eluru and Bhat, 2007;Eluru et al, 2008) models are among the models that do consider the ordered nature of crash severity. However, there is an increasing tendency towards using unordered response models, as well.…”

Evaluating the Effects of Road Geometry, Environment, and Traffic Volume on Rollover Crashes

“…The results showed that truck-driver behaviours, sharp curves, and turning manoeuvres were associated with higher rollover risk. Khattak and Rocha (2003) evaluated the rollover intensity and severity of driver injury of Sport Utility Vehicles (SUVs). A binary logit model was applied to investigate the likelihood of rollover occurrence.…”

The role of alternative fuel vehicles: Using behavioral and sensor data to model hierarchies in travel

“…Analysis of variance (ANOVA), Chi-square test and t-tests are the most commonly used methods comparing various groups (Subhashini and Arumugam, 1981;Simons-Morton et al, 2005). Ordinary Least Square (OLS) models including linear and logistic regressions are frequently applied to find the relationships between outcomes and associated factors (Khattak et al, 1995;McElroy, 1967;Dissanayake and Perera, 2011;Khattak and Rocha, 2003). Some studies have noted the hierarchical nature of behavioral data and applied multi-level models to explain relationships (Treno et al, 2003;Huang et al, 2008;Kim et al, 2007).…”