Understanding the Factors That Are Associated with Motorcycle Crash Severity in Rural and Urban Areas of Ghana

Table 4

Mixed multinomial logit model estimation results for urban motorcycle crashes.

Variable			Marginal effects
	Parameter estimate	t-statistic	Fatal injury	Incapacitating injury	Minor injury	No injury
Constant [FI]	1.448	2.67
Straight and flat road [FI]	−0.441	−3.14	−0.0488	0.0214	0.0205	0.0069
Rider age more than 60 [FI]	0.915	2.98	0.0015	−0.0007	−0.0006	−0.0002
Rider age between 45 and 60 [FI]	0.513	4.08	0.0056	−0.0024	−0.0023	−0.0009
Light rain condition [FI]	1.055	2.62	0.0011	−0.0003	−0.0005	−0.0002
Poor visibility [FI]	1.648	2.70	0.0007	−0.0003	−0.0003	−0.0001
Head-on collision [FI]	1.441	15.18	0.0436	−0.0184	−0.0169	−0.0083
Rider loss control [FI]	0.402	1.98	0.0016	−0.0007	−0.0007	−0.0002
Sideswipe [FI]	−0.544	−4.25	−0.0065	0.0027	0.0025	0.0012
Multi-vehicle collision [II]	−0.363	−4.08	−0.0297	0.0126	0.0116	0.0055
Constant [II]	2.057	3.95
Dark and unlit lighting condition [II]	−0.125	−0.53	−0.0025	0.0062	−0.0026	−0.0012
Standard deviation for “dark and unlit roadway” (normally distributed)	4.463	1.95
Gravel road [IHI]	0.315	1.70	−0.0005	0.0016	−0.0008	−0.0003
Constant [MI]	1.778	3.40
Collision at T junction [MI]	0.432	5.68	−0.0035	−0.0086	0.0152	−0.0031
Paved roadway [MI]	−0.310	−5.22	0.0094	0.0195	−0.0355	0.0066
Collision at signalized intersection [MI]	0.148	1.68	−0.0008	−0.002	0.0036	−0.0007
Collision with pedestrian [NI]	−5.437	−5.43	0.0002	0.0003	0.0003	−0.0008
Male rider [NI]	0.924	1.77	−0.0155	−0.0338	−0.0313	0.0805
Steering defect [NI]	0.866	3.54	−0.0005	−0.0009	−0.0008	0.0022
Multiple motorcycle defects [NI]	1.210	3.35	−0.0003	−0.0006	−0.0006	0.0015
Model statistics
Number of observations	6617
Log-likelihood restricted	−9173.11
Log-likelihood at convergence	−7848.12
McFadden pseudo-	0.14