Investigating Different Types of Red-Light Running Behaviors among Urban e-Bike Rider Mixed Groups

Table 5

Random parameter logit model estimation results (with interaction effects).

		NRLR	AEP	WRLR	AEP	IRLR	AEP
Rider attributes
E-bike type	Delivery e-bike	−0.25 (0.12)	−67%	0.43 (0.15)	28%	0.31 (0.11)	44%
Age	Older	—	38%	−0.45 (0.16)	−17%	−0.78 (0.24)	−34%
Behavior characteristics
Traffic light status	The middle stage of the red light	—	23%	−1.36 (0.29)	−23%	−1.67 (0.14)	−16%
	The late stage of the red light	1.43 (0.36)	53%	—	−15%	—	−15%
	Std. dev.	1.26 (0.29)	—	—	—	—	—
Waiting position	Unreasonable position	—	−32%	2.42 (0.31)	52%	3.43 (0.38)	35%
Signalized intersection characteristics
Traffic light type	Countdown signal	0.56 (0.21)	−5%	−0.31 (0.15)	3%	—	3%
Crossing length	—	—	9%	0.07 (0.02)	9%	−0.04 (0.01)	−12%
Traffic environment
The traffic volume	—	0.15 (0.12)	2%	—	2%	—	−4%
No. of individuals waiting upon arrival	—	—	−4%	—	3%	−0.12 (0.18)	−4%
Time of day	Afternoon	0.49 (0.23)	16%	—	−4%	—	−4%
Interaction term	e-bike type × age	—	34%	—	34%	−0.46 (0.22)	−21%
	e-bike type × time of day	0.62 (0.17)	58%	—	−12%	—	−12%
	e-bike type × waiting position	—	−6%	—	15%	0.43 (0.14)	15%

The symbols , , and indicate significance at 10%, 5%, and 1% levels, respectively.