Exploring the Application of the Linear Poisson Autoregressive Model for Analyzing the Dynamic Impact of Traffic Laws on Fatal Traffic Accident Frequency
Table 3
Fatal traffic accident analysis (standard errors in parentheses).
Model
ARIMAX (1, 1, 0)
PAR (2)
NB
NB with a time trend
Intercept
—
−6.821 (0.328)
−5.835 (0.347)
7.118 (0.29)
Gasoline price
3.836 (32.897)
−0.074 (0.028)
−0.077 (0.028)
−0.05 (0.02)
Beer consumption
575.665 (514.852)
2.37 (0.204)
1.805 (0.242)
0.393 (0.211)
Belt law
191.826 (66.103)
−0.157 (0.037)
−0.225 (0.048)
0.029 (0.039)
DUI toughened penalties
−98.515 (77.537)
−0.279 (0.043)
−0.445 (0.048)
−0.227 (0.042)
Red running
58.631 (63.765)
0.301 (0.059)
0.162 (0.059)
0.096 (0.039)
Alcohol law
−144.166 (67.999)
−0.219 (0.049)
−0.221 (0.06)
−0.228 (0.041)
Linear trend
—
—
—
−0.007 (0.003)
—
0.706 (0.093)
—
—
—
−0.194 (0.079)
—
—
—
—
0.006
0.002
N2
42
42
42
42
Log-likelihood
−237.85
−222.116
−254.633
−232.543
AIC
491.71
460.232
525.27
483.087
BIC
505.61
474.133
539.168
498.726
1The dispersion parameter in the NB model. 2N is the number of samples.
