 By Bruce Arnold
OPEN LETTER TO D. HOUSTON & L. RICHARDSON
Two Poli-Sci PhDs Caught Piling It High & Deep (Thanks to GG & HG) May 2008 DOWNLOAD THIS ARTICLE IN PDF FORMAT
David J. Houston, Ph.D.
...published in the November 2007 issue of the American Journal of Public Health, you claim:
"We examined cross-sectional time-series data from the 50 states and the District of Columbia for the period 1975 through 2004.... On average, when compared to state experience with no helmet mandate, universal helmet laws were associated with an 11.1% reduction in motorcyclist fatality rates, whereas rates in states with partial coverage statutes were not statistically different from those with no helmet law. Furthermore, in the states in which recent repeals of universal coverage have been instituted, the motorcyclist fatality rate increased by an average of 12.2% over what would have been expected had universal coverage been maintained. Motorcyclist safety has been compromised in the states that have repealed universal coverage and is likely to be compromised in other states that abandon these statutes." After paying $22.00 for the privilege of downloading a PDF copy of your "research" allegedly supporting these contentions, I found that it did not. In fact, I suspect informed readers would question whether your published methodology and analysis even qualifies as objective research: YES, I accept that a 51-by-30 cell grid would be a logical manner in which to organize 30 years of motorcycle fatality statistics for fifty states plus the District of Columbia. But NO, I do NOT accept that dividing motorcycle fatalities by motorcycle registrations to compute "Motorcycle Fatalities per 10000 Registrations" yields a reliable statistic to support objective calculations and comparisons across your 1530-cell state-year grid. As your fellow Ph.D. Margaret Sweeney clearly conveyed to her National Transportation Safety Board ("NTSB") superiors last September, our government's motorcycle registration statistics may be "seriously understated"... ...which means that any statistical measures based thereon may be materially misstated as well. Besides, a far more realistic basis for comparison than fatalities per registration would be fatalities per mile, but I guess even you two were wise enough to recognize that the motorcycle vehicle miles traveled ("VMT") statistics compiled by the National Highway Traffic Safety Administration ("NHTSA") are meaningless... ...as would be any computations or conclusions derived therefrom. But let's say, for the sake of discussion, that the "Motorcycle Fatalities per 10000 Registrations" statistics you computed were an objectively unbiased and reliably accurate basis for analysis and comparison. Well, according to the information summarized in Table 1 of your article... ...the mean values computed from your 30-year database for "Motorcycle fatalities per 10,000" are as follows:
In other words, before you introduced all your "additional variables" to the analysis, the data shows that state-years with a universal helmet law had the HIGHEST motorcycle fatality rate, while state-years with no helmet law had the LOWEST! I suspect that wasn't where you wanted your "scientific" research to conclude, because it certainly didn't. Instead, you proceeded to alter the equation by introducing more variables ... rational-sounding attributes like temperature, precipitation, age, income and alcohol consumption ... but each with creatively-defined if not subjectively-assigned values ... one after another until finally your computations supported your intended conclusions. And notably absent from your cherry-picking was any acknowledgement that in many more recent accidents riders are colliding with supersized SUVs and similar light truck vehicles ("LTVs") resulting in disproportionately high fatality-to-injury ratios, as Moonrider conveys so graphically in her two-part video on "Dangerous Designs": http://tinyurl.com/5qs9zq (Part 1) http://tinyurl.com/6or6uj (Part 2) You also made no allowance for the fact that "...in 2002 The Harvard University Center for Risk Analysis estimated that each year 1.5 million accidents, 560,000 injuries and 2,600 deaths were due to phone use in moving vehicles.[4] And there were only 86 million cell phone service subscribers in the U.S. in 2002.[5] As of May 2007, we now have more than 236 million cell phone subscribers nationwide ... almost three times as many ... and 73% of them are talking while they're driving![6]" We substantiated that here: I understand that professors must "publish or perish". I can further appreciate that since political science is your chosen domain, the statistics of spin might be your professional specialty. Going forward, however, I ask that you please ply your trade somewhere other than the motorcycle safety arena. You may have had good intentions, but like so many well intended do-gooders... ...you fail to appreciate the difference between the utility of helmets and the futility of helmet laws. And consequently, by shifting attention away from a much-needed focus on crash avoidance, you are costing far more motorcyclists' lives than can ever be saved by your myopic advocacy of a failed public policy based on crash survival. Speaking strictly for myself and no other individuals or organizations,
Bruce Arnold
DOWNLOAD THIS ARTICLE IN PDF FORMAT top of page series page home page biker stuff donate biker news LdrLongDistanceRider.com is a bikers' rights, motorcyclists' issues and long distance motorcycle riding resource for touring endurance riders and extreme cruising on bikes by Aprilia, BMW, Buell, Ducati, H-D/HD/Harley-Davidson, Honda, Kawasaki, Moto-Guzzi, Norton, Suzuki, Triumph, Vengeance, Victory, Yamaha and other makes.
|
In the abstract of your article "Motorcycle Safety and the Repeal of Universal Helmet Laws"...
|
||||
 |
||||
|
||||
|
||||
|
||||
|
||||
|
||||
Privacy Links Legal BBG4500
Copyright © 2005-2009 Bruce Arnold: Long Distance Motorcycle Riding & Bikers Rights
Miami Web Design IIHS Lid Laws