I would argue on several grounds that discounting the subjective perceptions and experiences of road users would be a mistake when considering what factors influence road safety.
In terms of accurately reflecting the reality of road use in London can the subjective opinions of road users be said to not provide any illumination in to their behaviour and interactions?
Can the results of even high quality trials ever be said to be free from the influence of the perceptions and agenda of those that commission, execute, analyse or report on those data?
Once such a trial is completed the use to which its results are put is unlikely to be rigorously objective. Will policy be applied that positively affects the reality on the road or positively affects the figures in the next analysis?
An acknowledgement that such human factors are at play would help support the analysis, diminish the risk of mis-interpretaion of the data and its misuse.
Ben Lovejoy said:
Policy needs to be determined by reality; perceptions then need to be addressed by publicity and education.
Policy, reality, perceptions, publicity/education are all such interrelating factors on peoples behaviour and each other that attempting to formulate their application from a statistical analysis while discounting any psychological factors would be like jellying a wall to a nail.
Due to the complexities involved and the fast changing nature of our society, possibly unfortunately, the effective delivery of policy that will deliver a reduction in death, injury and fear and maintain or improve mobility on the roads will probably require a balance of intuition and ideology as well as an informed and developing knowledge of transportation planning.