In the early 1990s, data from over 3000 Dutch drivers were analysed, to establish relationships between possible key variables in drink-driving research. The data came from four samples of drivers, participating in two consecutive experiments on enforcement in a western part of Holland. Drivers were stopped at random, interviewed, and tested for alcohol. They were asked three questions about driver experience and perception of police enforcement, and another question about awareness of police enforcement due to publicity. The data were analysed by a version of the BMDP(4F) statistical software package. Chi-square tests were applied to assess the statistical significance of various effects, and in relation to various statistical models, including loglinear models. Analyses were also made of standardised residual. The proportion of drivers over the legal limit, a blood alcohol content (BAC) of 0.05%, decreased during the study period. However, the results of the loglinear analysis do not indicate that drivers' behaviour was directly affected either by the publicity or by what they actually noticed. On the other hand, drivers may make a rather general assessment of the driving while intoxicated (DWI) enforcement, which may have some importance.
Abstract