Exploring the factors affecting motorway accident severity in England using the generalised ordered logistic regression model
journal contributionposted on 22.01.2016 by Paraskevi Michalaki, Mohammed Quddus, D.E. Pitfield, Andrew Huetson
Any type of content formally published in an academic journal, usually following a peer-review process.
Problem The severity of motorway accidents that occurred on the hard shoulder (HS) is higher than for the main carriageway (MC). This paper compares and contrasts the most important factors affecting the severity of HS and MC accidents on motorways in England. Method Using police reported accident data, the accidents that occurred on motorways in England are grouped into two categories (i.e., HS and MC) according to the location. A generalized ordered logistic regression model is then applied to identify the factors affecting the severity of HS and MC accidents on motorways. The factors examined include accident and vehicle characteristics, traffic and environment conditions, as well as other behavioral factors. Results Results suggest that the factors positively affecting the severity include: number of vehicles involved in the accident, peak-hour traffic time, and low visibility. Differences between HS and MC accidents are identified, with the most important being the involvement of heavy goods vehicles (HGVs) and driver fatigue, which are found to be more crucial in increasing the severity of HS accidents. Practical applications Measures to increase awareness of HGV drivers regarding the risk of fatigue when driving on motorways, and especially the nearside lane, should be taken by the stakeholders.
This research was undertaken as part of an Engineering Doctorate project jointly funded by the Centre of Innovative and Collaborative Construction Engineering (CICE) at Loughborough University and Balfour Beatty. The support of the Engineering and Physical Sciences Research Council is gratefully acknowledged (EPSRC Grant EP/F037272/1).
- Architecture, Building and Civil Engineering