Harold J. Bodeker - Comment

I would like to applaud the NHTSA for conducting research on the effectiveness of motorcycle rider training on motorcycle crashes. However, I have a concern with the methodology of relying on voluntary responses. In the NHTSA's revue of driver education curricula (Feasibility Study of Evaluating Driver Education Curriculum - DOT HS 811 108 - April, 2009), the authors reviewed the DeKalb Study among others. In reviewing the DeKalb study's comparison of the driver education groups and the control group, the authors came to the conclusion that "...these comparisons do not provide a valid test of driver education, because students were self selected into the analysis...". If the NHTSA's own researchers concluded that self selection by study subjects rendered the resulting comparisons invalid, then the same invalidity would taint any comparisons of trained versus untrained motorcycle riders if they self selected into the study as volunteers. I encourage the NHTSA to proceed with this study, but strongly encourage a more scientific methodology (i.e. random sampling) be used to collect data from the study subjects. Otherwise, there is the risk that the study outcome, whatever it is, may be attacted as invalid, just as the DeKalb study was described by the NHTSA's own reasearchers.