Research Article

Differences in Rental and Nonrental Car Crashes

Table 3

Estimation results of binary logit, rare event logit, and Firth logit models.

VariableBinary logistic modelRare event logistic modelFirth logistic model
Coeff.Std. err. valueCoeff.Std. err. valueCoeff.Std. err. value

Driver demographics
Male0.99860.0377<0.0010.99810.0377<0.0010.99810.0376<0.001
Age (25–44)−0.92390.0356<0.001−0.92410.0356<0.001−0.92410.0356<0.001
Age (45–64)−1.32820.0400<0.001−1.31810.0409<0.001−1.31810.0409<0.001
Age (65+)−2.22220.111<0.001−2.21700.1107<0.001−2.21710.1104<0.001
Driver behaviors
Inattention0.17520.0284<0.0010.17510.0284<0.0010.17510.0284<0.001
Impaired driving−0.10840.05140.035−0.10760.05150.037−0.10760.05140.036
Poor driving0.35670.0657<0.0010.35830.06537<0.0010.35830.0658<0.001
Poor handling0.40390.11920.0010.40970.11830.0010.40970.11880.001
Aggressive driving0.50400.15950.0020.51460.15750.0010.51460.15860.001
Other driver factors0.22090.08480.0090.22360.08790.0080.22360.08460.008
Control variables
No vehicle factor−0.00820.08460.9230.00520.08450.9500.00530.08450.950
No road factor−0.09180.06640.167−0.09350.06610.157−0.09340.06640.159
No environmental factor−0.10590.17390.543−0.11880.17340.493−0.11860.17280.492
Constant−3.17480.1968<0.001−3.15650.1977<0.001−3.15660.1957<0.001

Note.   and   denote statistical significance at 95%  and  99% confidence levels.