Investigating the Differences of Single-Vehicle and Multivehicle Accident Probability Using Mixed Logit Model

Table 2

The estimated results of mixed logit model with three crash categories.

Variable	Estimated parameter	-statistic
Constant [SV]
Constant [MV]
Temporal characteristic
Weekend [MV]
November [MV]
4 a.m.-5 a.m. [MV]
Traffic characteristic
Low speed limit indicator [MV]
Speed gap [SV]
Speed gap [MV]
Hourly traffic volume [MV]
Truck percentage [MV]
Road characteristic
Number of entering ramps per lane per mile [MV]
Curvature [MV]
Length of segment [SV]
Length of segment [MV]
Long remaining service life of rutting [MV]
Environmental and road surface characteristic
Visibility [SV]
Wet road surface [SV]
Wet road surface [MV]
Chemical wet road surface [MV]
Random parameters
Mean of log of hourly traffic volume [MV]
Std. dev. of log of hourly traffic volume; (normal distribution)
Mean of inside shoulder width [MV]
Std. dev. of inside shoulder width; (normal distribution)
Mean of wet road surface [MV]
Std. dev. of wet road surface; (normal distribution)
Fit statistics
likelihood
AIC (smaller is better)
BIC (smaller is better)