Monday, October 19, 2009

More details on Crash Causation Study

We've launched our Crash Study Petition. Please sign it.

We promised more information on the states that have supported the Pooled Study initiative, details from FHA are here. The states that have already chipped in a total of $560K are: New York, Ohio, Oklahoma, Rhode Island, Texas and Wisconsin. Good work, states, and if any readers feel like contacting their state's Motorcycle Safety Officer, the contact details are in our petition resource page. Or you could talk to your state DOT or your state reps about it.

According to the SAFETEA-LU fact page, $2814K of the $3187K has been paid out.

Our best estimate of the study costs, based on informed sources, is that Dynamic Science is charging between $8000 and $9000 per crash. That would make their part of the 300 crash study worth something like $2.55M, and OSU's maybe $637K, each give ot take 100K. Universities customarily keep a percentage of grant money for overhead and salaries, this percentage looks about usual. Dynamic Science has three researchers in South California, who did the pilot study starting in December 2008 and finished this spring. They are scheduled to start on the rest of the initial 300 crash studies soon.

The blog for Feb 1, 2009 has a lot of background info. It details how Senator James Inhofe, R-Okla, rewrote the crash study to award the contract for the study to University of Oklahoma as pork. Dr. Samir Ahmed, reportedly, was unhappy at being stumped with the project. The OSU Transportation Center is known for road infrastructure research.

The funding for the study went through the Federal Highway Administration, not the National Highway Traffic Safety Admin., which usually does motorcycle-related studies. The hand of the AMA (American Motorcyclist's Assoc.) can be seen in this odd decision. NHTSA had developed (deservedly) a reputation among bike groups of slanting their research to make political points about helmet use, leading to distrust of NHTSA research findings. The use of Federal Highway Admin bypassed this problem.

In fact, there is very little chance that this study will be used in the helmet wars. AMA has been policing the issue and as the study is to be done in California, a helmet state, there is not likely to be any data about the effects of crashes without helmets. Informed sources suggest that the researchers don't consider helmet use to be a significant cause of crashes, so it is just not part of their brief to consider this factor. In practical terms, unless it is proposed to collect data in some other state as well as California, we are safe from the helmet issue becoming a factor. In addition the OECD research methodology was developed for the Maids study initially and helmet use is not an issue in Europe. Looks like AMA has done a good job in keeping this study clean of the helmet issue, so nobody need fear that it will come up.

So far, the other research problem has not cropped up yet, no-one has flown the insurance industry's pet issue of mandatory speed governers, hopefully we won't see this either. Speed governers would be counter productive, anyone who wants to hop up their engine computers can bypass the governer, and we are pretty sure that modified bikes are a major safety issue, again from Maids. What would happen, if a mandatory say 120 MPH speed governer law was enacted, would be that the bike manufacturers would start putting in 120 MPH brakes, tires etc., so when bikers swapped out the chips for 200 MPH chips, the 120 MPH components, especially tires, would fail in huge numbers, giving us a net safety downgrade. So the speed governer issue is a non-starter, even though nobody expects the new study to conclude that speed in not a factor in crashes, as Hurt did.

If the numbers can just be got up to 900 somehow, we'll get a good study, free of at least the usual biases. Judging by what happened after the armed forces did some studies of their members crashes, it's more likely that training will be the big issue.

Upcoming: we'll look into the OECD methodology, with Maids as an example, and see how these studies work.


Post a Comment

Links to this post:

Create a Link

<< Home