This study analyses data on the use of Daytime Running Lights (DRL) measured in the Netherlands over the months of November 1989-November 1991, and May and June 1992. By means of this analysis, the study investigates the usefulness and reliability of models to estimate the use of DRL for the months and locations in which no measurements are performed, such as the December 1991-April 1992 period. The ultimate aim of these models is twofold: a) to evaluate publicity campaigns, for instance the Dutch national publicity campaign started on 6 November 1992; and b) to evaluate the effect of DRL on traffic safety. The main results of the study show that: (1) the logistic model seems to provide sufficient possibilities for estimating the use of DRL, with the exception of the winter months; (2) when a trend-break occurs in the use of DRL caused by the publicity campaign, the (logistic) model cannot be used for the period after this trend-break; (3) for the benefit of a statistically justified accident analysis, a 'reasonable' reconstruction of the use of DRL seems possible; and (4) the existing computer program is not capable of making statistically reliable calculations using the estimation model yet.
Abstract