Abstract
It is shown how the method of maximum likelihood in connection with a bivariate negative binomial distribution may be used to separate treatment effects from regression to the mean effects when analyzing data giving the frequency of accidents in two time periods for a number of sites on a road network. The method is illustrated using data from the Lothian Road Accident Statistics System (LORASS).
| Original language | English |
|---|---|
| Pages (from-to) | 168-169 |
| Number of pages | 2 |
| Journal | Traffic Engineering and Control |
| Volume | 29 |
| Issue number | 3 |
| Publication status | Published - Mar 1988 |
| Externally published | Yes |