Abstract
A practical resolution of the low mean value problem which is a barrier to goodness of fit testing for the commonly used generalised linear accident model, and the mechanism of these models were discussed. The number of loss-of-control injury accidents, over 5 years, against 24 h vehicle flow for 392 approaches to intersections in New Zealand, were presented. Results showed that the accident variance grew larger than the accident mean as the mean increasesed. This indicated that a negative binomial distribution might be appropriate for the number of accidents at a given flow.
| Original language | English |
|---|---|
| Pages (from-to) | 417-427 |
| Number of pages | 11 |
| Journal | Accident Analysis and Prevention |
| Volume | 34 |
| Issue number | 4 |
| DOIs | |
| Publication status | Published - Jul 2002 |
Keywords
- Generalised linear model
- Low mean value
- Negative binomial
- Pearson's X
- Poisson
- Scaled deviance