Investigation of Contributing Factors to Traffic Crashes and Violations: A Random Parameter Multinomial Logit Approach

Table 4

Estimation results of the random parameters multinomial logit model for the seriousness of traffic violations.

Characteristic	Minor violation		Serious violation		Crash with violations
	Parameter estimate	Standard deviation	Parameter estimate	Standard deviation	Parameter estimate	Standard deviation
Constant	−0.368	0.033	2.600	0.041	2.423	0.574
(Std. dev. constant)					1.238	0.557
Gender
Female	0.446	0.057	−0.869	0.046	−0.917	0.487
(Std. dev. female)					0.773	0.329
Race/ethnicity
African American			1.537	0.391
(Std. dev. African American)			3.013	0.565
Hispanic	0.112	0.044
Time period factors
Afternoon peak (4–6 p.m.)	0.343	0.068
Evening (6–10 p.m.)					0.519	0.105
Seasonal factors
Fall			−0.207	0.042
Commute mode
The percentage of driving to work of the driver’s residence area (ZIP)	0.151	0.025	0.228	0.019
Commute travel time
Mean travel time to work of the driver’s residence area	0.074	0.026	0.043	0.020
Household income
Median household income of the driver’s residence area	−0.241	0.027

Note., , and indicate significance at 10%, 5%, and 1% levels, respectively. Std. dev.: standard deviation.