Research Article
Distinguishing between Rural and Urban Road Segment Traffic Safety Based on Zero-Inflated Negative Binomial Regression Models
Table 4
Parameter estimates of ZINB model for total crash frequency.
| | Parameter | DF | Estimate | Standard error | t value | P value |
| | Negative binomial regression part |
| | Intercept | 1 | 0.850 | 0.161 | 5.29 | <0.0001 | | Numberofla (2 lane) | 0 | — | — | — | — | | Numberofla (4 lane) | 1 | −0.438 | 0.107 | −4.10 | <0.0001 | | RoU (rural) | 1 | −0.270 | 0.095 | −2.84 | 0.0045 | | RoU (urban) | 0 | — | — | — | — | | ADT_1000 | 1 | 0.075 | 0.004 | 18.20 | <0.0001 | | LENGTH | 1 | 0.395 | 0.070 | 5.61 | <0.0001 | | FFS | 1 | −0.011 | 0.004 | −2.99 | 0.0028 |
| | Logistic regression part |
| | Inf_Intercept | 1 | 8.214 | 0.913 | 8.99 | <0.0001 | | Inf_ADT | 1 | −0.028 | 0.013 | −2.20 | 0.0277 | | Inf_LENGTH | 1 | −71.713 | 8.739 | −8.21 | <0.0001 | | _Alpha | 1 | 1.435 | 0.065 | 22.20 | <0.0001 |
|
|
