ABSTRACT
This chapter discusses the number of different ways of calculating accident record regarding two features of importance; how 'safety' is conceptualized when a certain method is used, and what the likely consequences are of using it. It discusses a few approaches to the collision criterion that do not fit the broad classification made above with regards to exposure. Calendar time in relation to accidents has apparently attracted interest from researchers in traffic safety, in terms of often being used to keep exposure somewhat constant. The most basic and straightforward type of grouping in traffic safety is drivers with collisions versus drivers without collisions for a specified period. The use of case-control also has interesting methodological consequences that are similar to the ones encountered for the other methods described here. In the end, what cases and controls you chose is a mirror image of how you conceptualize safety.
