Page 80 - tmp_Neat

The remainder of this chapter is organised as follows. In the next section, the data used in this

research and the estimation model are presented. The results are discussed in Section 5.3, and
the last section provides the concluding remarks.

5.2 Methods

5.2.1 Data

The primary data used in this study were provided by VicRoads and contained information on

all police-reported collisions in Victoria, Australia. It included information on environmental,
temporal, road user, roadway and vehicle characteristics. This study focuses on two-vehicle

angle collisions involving at least one heavy vehicle. Heavy vehicles in this research include

vehicles in several categories in the VicRoads classification scheme: prime mover only,
prime mover (single trailer), prime mover (B-double and B-triple), prime mover (number of

trailers unknown), light commercial vehicle (rigid<= 4.5 tonnes), heavy vehicle (rigid > 4.5
tonnes), and rigid truck (weight unknown). Note that buses and other types of heavy vehicles

(machinery, farm equipment, etc.) were not included in this study.

According to the VicRoads classification, crash injury severity is classified into four
categories: fatal injury (killed or died within 30 days; serious injury (sent to hospital, possibly

admitted); other injury (typically requires medical treatment); and non-injury. From 2006 to

2016, 3675 people were involved in two-vehicle angle collisions involving at least one heavy
vehicle in Victoria, Australia. Of the 3675 persons involved, 107 (2.9%) were fatally injured,

759 (20.7%) were seriously injured, 1085 (29.5%) had minor injuries, and 1724 (46.9%) had
no injury. In this study, fatal and serious injuries were combined into a severe injury category

because of the very small number of fatal injuries. In addition, other injury and non-injury
were combined into a minor injury category because of their substantially lower average

social cost. In the final sample, severe injury comprised 23.6% and minor injury comprised

76.4%.

In addition to the crash data from VicRoads, information on traffic volume and road features

(e.g. road division marking) were extracted from the Australian Urban Research
Infrastructure Network (AURIN). The variables used in estimating the final model are
63