The University of Adelaide CENTRE FOR AUTOMOTIVE SAFETY RESEARCH

home   /   centre for automotive safety research   /   Publications   /   List   /   Details

Publication Details

TitleControl groups, randomisation, double blinding, meta-analysis: Can road safety research learn from evidence based medicine and social welfare?
AuthorsHutchinson TP
Year2004
TypeConference Paper
AbstractThe practice of medicine, these days, is supposed to be evidence based. Important features of this are the use of randomised experimentation to compare treatment and control interventions, and the systematic reviewing of such research using metaanalysis. The idea has spread to social welfare, criminology, and education. The present paper considers whether evaluation of road safety interventions should use similar methods, or whether evidence-based everything is a fad that we can safely ignore, impracticable in road safety. A real-world road safety intervention is typically a much more complex process than, say, a comparison of aspirin with placebo. However, some of the experiments in social welfare, criminology, and education must have been just as difficult as those in road safety. There is much in the statistical, medical, and social welfare literatures on the desirability of a control group, randomisation, and double-blinding, and the dangers of merely making a before-after comparison, of allocating experimental units to groups in a non-random manner, and of allowing either the experimental units or the researchers to know which is in which group. However, there are many issues in transport safety for which the disadvantages of a rigorous methodology will outweigh the advantages. Nevertheless, for some issues, randomisation (and other precautions) ought to receive more consideration than at present. If that happens, planning the monitoring of a road safety intervention will take longer and need more resources.
ISBN07307 24921
Conference NameRoad Safety Research, Policing and Education Conference
Conference AbbreviationRSR
Conference LocationPerth, Australia
Conference Date14-16 November 2004
Page Count10

Reference
Hutchinson TP (2004). Control groups, randomisation, double blinding, meta-analysis: Can road safety research learn from evidence based medicine and social welfare?. Road Safety Research, Policing and Education Conference, Perth, Australia, 14-16 November 2004.


Files Available for Download
CASRcontrolgroups470.pdfscanned PDF, OCR'ed