Originally Posted by bwileyr
Bruce is correct to qualify this statement with the modifier "possible". All things are possible but the information provided on the cited URL provides so little data that no credible analysis of correlation between Bike Ed Hawaii and a "lower accident rate" can be made.
Accident per registration is a neat statistical trick when discussing risk. So is "involvement rate."
Even if accepting the assumption that bicycle registration is a correct indicator of the number of cyclists, it is no indicator of number of miles or time spent cycling or in what kind of environment. Who knows who registers bicycles in the various counties of Hawaii or if registration is remotely related to time or miles cycling?
Accidents are not defined at all, accident severities are unknown; does a bent rim or skinned knee equal a crippling injury or permanent disability when computing the so-called "involvement rate?"
More to the point there is no data presented that the Bike Ed students in Honolulu had any different "involvement rate" than the majority of the cyclists from Honolulu who have never had this training. The population density, geographic differences and cycling environment between Honolulu and the rest of Hawaii are so different that comparing any reported difference in the vague term "involvement rate" requires more careful analysis than just tossing out numbers. For the last reported year,1989, Honolulu has one tenth the land mass of Hawaii, 75% of the population, and 77% of the "accidents" whatever those are in this shaky listing. Given the lack of data, the "safety record" of the formally trained cyclists in Hawaii in comparison to untrained cyclists is anybody's guess.