In a number of field studies in which the conflict-observation method was used there was some doubt how long the observation should be continued. Two main problems were analyzed; (1) how long should an observation be made before casual as well as systematic fluctuations can be eliminated as much as possible (representative measuring period), and (2) how long should a period be to get a fair idea of the situation if the conflict observations are being carried out to make a diagnosis or to determine the effects of countermeasures in before and after studies. Specific questions also play a role. There is a difference depending on whether all conflicts at a junction or only driving to the left manoevres are to be evaluated. Methods used in different countries are discussed.
Samenvatting