Page 84 - tmp_Neat

Number of lanes

1-2 (reference) 68.6 63.9

3-4 29.3 34.5

5-6 2.0 1.6

Road division marking

Divided double 9.4 11.4

Single divided (reference) 13.5 13.5

Not divided 44.0 43.6

Others 33.1 31.5

In addition to the variables shown in Table 5.1, several other variables were included in the
preliminary analyses but were found to be statistically insignificant at the 90% confidence

level. The insignificant variables included the state where the driver's license was issued, road

surface and weather conditions, season of the year, day of the week, number of people
involved in the collision, light conditions, special road facilities, type of traffic control

device, vehicle weight, road shoulder width, percentage of heavy vehicles and average annual
daily traffic volume (AADT).

5.2.2 Binary Logit Model

A summary of existing studies on the binary logit model is presented in Section 2.2.2.1 of

Chapter 2. The dependent variable in this study is the injury severity of vehicle occupants

involved in two-vehicle angle crashes that include at least one heavy vehicle. It is classified
into two categories in this research: severe injury and minor injury. Considering the nature of

the dependent variable, the use of the binary logit regression model is considered to be
appropriate. This model has been widely applied by previous researchers in road safety for

modelling a dependent variable with a dichotomous outcome (Anowar et al., 2013; Johnson

et al., 2011; Rifaat et al., 2009, 2011; Tay et al., 2008, 2009; Tay & Choi, 2016; Weng &
Meng, 2014).

67