A method of Driver Performance Measurement was developed using a content validity and systems psychology basis. It showed very high reliability. For evaluation purposes two new groups of observers were trained by the original team of experts. Extensive statistical analysis of results showed that their ratings were of similar very high reliability and validity. This method is applicable to research where the conventional large group, long duration type of research in terms of accident records cannot be validly used. The method can be used for research comparing small groups to improve driver education or driver licence examining methods. Reliable and valid results cannot be expected unless the measurements are made by observer/raters with proper background who have had special training of the type which has been demonstrated and evaluated.
Abstract