This report provides the results of the further analysis of the second phase of the research project "Risk judgements and decisions of traffic participants". In that second phase an experiment was done, in which young bicyclists (12-16 years old) were interrogated in different ways about their behaviour in the approach of an intersection, when they have the intention to turn left on that intersection. The interrogation was also aimed at obtaining insight into the factors and arguments which could influence the behaviour concerned. The interrogated youngsters were selected on the basis of the number of observations of their actual behaviour, which was available for each of them. Several times, namely, unobtrusive video registrations were made at one specific intersection of the approach behaviour of bicyclists turning left. The results of the experiment showed a reasonably large degree of consistency of observed behaviour in comparable circumstances, and also a reasonably large degree of consistency between two of the three different ways of interrogation. These two methods predicted (excepted for looking backwards) the behaviour studied reasonably. There was, however a systematic overestimation of active behaviour. The third method did not predict the studied behaviour at all. Some time later the youngsters were once more interrogated about their own behaviour in two other ways. This time in a more direct way and with another payment structure. Again with exception of looking backwards this time the two methods predicted the observed behaviour well in a relative as well as in an absolute sense. (A)
Abstract