. 2020 Oct 22;22(11):1191. doi: 10.3390/e22111191

Table 3.

Results of the mixed logit model.

Variable		Coeff. *	Std. Error	t	p-Value **
traffic control	signallight	−0.5801	0.0644	−9.01	<0.0001
traffic control	stopsign	0.2037	0.0587	3.47	0.0005
location	Offroad *^a	−2.5980 (2.9134)	0.5779	−4.50	<0.0001(<0.0001)
weather	Rain *^a	−1.2411 (1.7131)	0.3592	−3.45	0.0006(<0.0001)
	sleet	−0.8620	0.2479	−3.48	0.0005
	snow	−1.4201	0.2880	−4.93	<0.0001
	severcrosswinds	−0.7184	0.3243	−2.22	0.0267
road alignment	straigrade	0.1569	0.0488	3.21	0.0013
	curlevel	0.2380	0.0669	3.56	0.0004
	curgrade	0.3068	0.0800	3.83	0.0001
functional system	uinterstatehighway	−0.7046	0.0618	−11.40	<0.0001
functional system	uotherprincipalarterial	−0.5464	0.0480	−11.39	<0.0001
curb left	curbleft	−0.6472	0.0581	−11.13	<0.0001
median	positivebarrier	−0.5434	0.0548	−9.91	<0.0001
	Unprotected *^a	−1.8760 (−2.2862)	0.3472	−5.40	<0.0001(<0.0001)
	onewaypair	−0.6444	0.3111	−2.07	0.0383
factors	drug	2.0215	0.1466	13.79	<0.0001
factors	fatigue	0.7642	0.0935	8.18	<0.0001
Intercept		−2.1065	0.0278	−75.85	<0.0001
Log likelihood	−19,352 (df = 22)
Sample size	85,184

* Parentheses indicate the standard deviations of the random coefficient estimates. ** Parentheses indicate the p-value of the standard deviations of the random coefficient estimates. *^a Denotes random parameters.