Samenvatting
Four methods of obtaining information on driver behaviour are described. Reasons are given for preferences for the check-list method over those in which the driver gives a running commentary on what he sees and how he reacts to it and/or and observer gives a running commentary on the drivers behaviour.