Abstract
The paper deals with the evaluation of road safety counter-measures and the constraints imposed by lack of suitable data. It also points to how these limitations may be overcome through careful selection of methods of analysis. The use of control groups and collection of additional information such as exposure data are discussed.