Journal of Advanced Transportation

Research Article

Evaluation of Roundabout Safety Performance through Surrogate Safety Measures from Microsimulation

Table 8

Parameter estimates for crash models based on AIMSUN simulated conflicts and goodness-of-fit.


Parameter	quasi-Poisson model^∗			Negative Binomial model^∗
Parameter	estimate (s.e.)	t	t pr.	estimate (s.e.)	t	t pr.

β₁	0.859 (0.222)	3.88	<.001	1.117 (0.242)	4.62	<.001

β₂	5.77 (1.09)	5.31	<.001	6.38 (1.17)	5.45	<.001

β₃	2.299 (0.533)	4.31	<.001	2.397 (0.578)	4.15	0.003

	- 0.0034	-	-	0.2987	-	-

	0.4364	-	-	0.5964	-	-

	0.6430	-	-	1.4217	-	-

	0.727	-	-	1.607	-	-

	0.0034	-	-	0.2987	-	-

	0.1613	-	-	0.3299	-	-

	0.72	-	-	0.67	-	-

	100%	-	-	100%	-	-

Note: N is the data sample size, and is the fitted value of y_i, which is the actual measurement; is the mean value of the fitted values, while is the mean value of the actual measurements; dof stands for degree of freedom; stands for Pearson product moment correlation coefficient. (^∗) Note that in GenStat the dispersion parameter (fixed or estimated) is used when calculating standard errors and standardized residuals. In models with the Poisson and negative binomial, as well as geometric and exponential distributions, the dispersion should be fixed at 1 unless a heterogeneity parameter is to be estimated.