On Saturday, March 31, 2018 at 1:24:59 AM UTC+1, jbeattie wrote:

I tried to pull the study on LEXIS, but it is not published in a journal yet. I think the study just looked at accident rates pre and post-law. It apparently evaluated the success of the law rather than the success of conspicuous clothing.

So what Franki-boy is telling us is that Italians are scofflaws. Much less offensive, stereotyped opinions than that one can get you fired from even a tenured position at American colleges these days. Here's a case in your own backyard: https://www.nationalreview.com/2018/...ias-tribunals/

I don't know how you could possibly control one of these experiments.

No problem. In a real experiment, you would compare like with like, for instance you would establish that the law was obeyed or enforced where it was ignored, and that the mandated or available clothing was actually visible at night, and then you would measure the change in cycling rates over the period you're studying, and find a control group (presumably in a country next door with similar cycling conditions but without the visibility law), at which point, after you've made adjustments for all the variables you found, you should be able to make a comparison for your population of interest between two time periods within calculable margins of confidence. All of this sounds like very much more trouble and expense than was undertaken in that Italian academic's makework "study". In fact, her study, and Krygowski's slimy attempt to pass it off as meaningful, both remind me of the Zimmerman "study" of 76 "scientists" on which the lie that "97% of scientist agree that global warming is manmade" was based.

However, you mustn't go into the project with unrealistic expectations. This business, common in cycling and government circles, of a hard number to answer such complicated statistical questions exists strictly in the minds of those entirely ignorant of real-life demography and its representation in sampling statistics. I'm looking at you, Franki-boy. Statisticians, sociologists, psychologists, economists and other applied mathematicians who work in these fields -- and who have vastly more financial and human resources available than some random academic more interested in publication-for-promotion than knowledge--soon learn to temper their expectations of results to their understanding that correct formulation of the hypothesis and intelligent interpretation of the situation and the results are together more than half the requirement for formulation of a meaningful policy directive from sampling research.

AJ