Table 3.

Binary logit and heteroscedastic logit model results for crashes between motorcycles.

Variable	Binary Logit Model		Mixed Logit Model
	Coefficient Estimate	p-Value	Coefficient Estimate (Standard Deviation)	p-Value
Intercept	0.209	<0.001	−9.705	<0.001
Crash and environment characteristics
Midnight (00–06)	0.913	<0.001	−0.526 (1.795)	0.3939 (n.s.) (<0.001)
Dark without streetlight	0.820	0.0083	0.839	0.0080
Rural	0.552	<0.001	0.561	<0.001
Speed limit > 50 km/h	0.872	<0.001	0.885	<0.001
Separation-Traffic Island	0.249	0.0257	0.256	0.0235
Pavement surface-dry	0.456	0.0101	0.464	0.0094
Crash type
Head-on collision	1.032	<0.001	1.035	<0.001
Swipe collision—same direction	0.393	0.0042	0.397	0.0041
Right angle collision	0.496	<0.001	0.501	<0.001
Side impact collision	0.275	0.0096	−0.168 (0.985)	0.4697 (n.s.) (<0.001)
Motorcyclist characteristics
Male	0.524	<0.001	0.528	<0.001
Aged (55 years old+)	1.573	<0.001	1.358 (0.688)	<0.001 (0.011)
Without wearing helmet	1.895	<0.001	1.946	<0.001
LHM (250 cc+)	1.162	<0.001	1.181	<0.001
Motorcyclist’s BAC (+<0.3%)	1.823	<0.001	1.853	<0.001
Motorcyclist’s BAC (≥0.3%)	1.650	<0.001	1.686	<0.001
Motorcyclist’s BAC unknown	1.952	<0.001	1.995	<0.001
Counterpart motorcyclist characteristics
LHM (250 cc+)	1.422	<0.001	1.448	<0.001
Without valid driving license	0.588	<0.001	0.597	<0.001
Speed related violation	0.788	<0.001	0.794	<0.001
Distracted	0.464	<0.001	0.466	<0.001
Right of way violation	0.376	<0.001	0.376	<0.001
Motorcyclist’s BAC (+<0.3%)	0.994	<0.001	1.022	<0.001
Motorcyclist’s BAC (≥0.3%)	0.758	<0.001	0.774	<0.001
Model statistics
LL0			−5095
LL	−4470.251		−4465
R2	0.123		0.124
AIC	8991		8985

Note: n.s.: not significant at the 95% confidence interval.