Self-report is a valuable methodology of social research, especially in regard to offending behaviour on and off the road. While assessing the validity of self-report is rarely easy, taking objective measures of drivers' speed choice behaviour does provide an opportunity to assess the correlation with subjective measures. Noting studies where this has been possible, research carried out by the author and colleagues in England which allowed similar assessments is discussed. Significant correlations resulted between reported and observed speed in all six surveys comprising the research, but they were of a low order. Analysis of the data showed a consistent and systematic bias towards drivers travelling above the 30 mph speed limit reporting their normal speeds as lower than those observed, with those travelling below the 30 mph speed limit reporting their speeds as higher than those observed. These patterns are discussed in light of other studies which have found tendencies either to overstate or to understate, and the possibilities that drivers either have difficulty reporting accurately or choose not to do so. Psychophysical limitations on accuracy of self-report and the social and cognitive processes that could mediate them are considered. In particular, the notion that drivers may seek to comply with normative speeds, and that these can be perceived as safer than complying with posted limits are discussed in light of normal police practice to tolerate low margins of excess speed. (Author/publisher).
Abstract