Analysis of the Severity of Accidents on Rural Roads Using Statistical and Artificial Neural Network Methods

Table 6

The estimation of logistic regression model variables in the first step.

Variables	Factors	B	Std. error (SE)	Exp (B)	Wald	Sig	Exp (B) (95% CI)
Collision with	Motorcycle = 1	−4.529	2.085	0.011	4.718	0.030	0.0002–0.64
Type of collision	Head on = 1	−0.902	0.510	0.406	3.124	0.077	0.149–1.103
Type of collision	Rear-end = 3	−1.111	0.401	0.329	7.693	0.006	0.150–0.72
The type of maneuver of the guilty vehicle	Moving forward = 1	1.580	0.954	4.855	2.738	0.098	0.747–31.5
The type of maneuver of the guilty vehicle	Turning to left = 2	1.990	1.120	7.313	3.156	0.076	0.815–65.69
The type of maneuver of the guilty vehicle	Turning to right = 3	1.779	1.075	5.924	2.739	0.098	0.72–48.67
The type of maneuver of the guilty vehicle	Overtaking = 5	2.363	1.205	10.623	3.843	0.050	1.001–112.73
Road factor	The presence of obstacles and bumps = 4	2.310	0.901	10.07	6.568	0.010	1.72–58.97