Dear Frank, You have done me the honour of at least reading my attachment, and in your reply I will gloss over the comment where you seem to imply that I am deliberately telling lies.
A significant part of my note is reporting the work of others such as our former Government agency, the Transport Research Laboratory, and the report (with many references) by the Royal Society for the Prevention of Accidents. You are quite right that it may be better to steer clear of road accidents and predictions, which are difficult to analyse or prove. Perhaps concentrate on the other important topic, linked to global warming, the reduction of demand for electricity. Here your colleagues in the University of Cambridge have made a detailed analysis and some predictions too. If you disagree with any of the analyses or methods (and for the above reports) you could perhaps take issue with them. Returning to DST and Single Double Summer Time, the key factor is the shift away from purely ‘rural’ time (before artificial light) of a day centred on noon, to the current life styles where the middle of the effective day has moved to something like 3pm. I had hoped that my note was objective enough, and with sufficient detail, for readers to draw their own conclusions or even go back to the source documents. Regards, Doug > On 23 Nov 2016, at 16:09, Frank King <f...@cl.cam.ac.uk> wrote: > > Dear Doug, > > I read your message and your attachment. > > There are so many flaws in your analysis > that it is difficult to know where to > begin. > > Since you find road accident statistics > so compelling I shall confine myself to > demolishing all you have to say there... > > There is a wonderful book by Darrell Huff > "How to lie with statistics". You seem > to have studied this carefully. > > Still, since you want us all to tell > lies about the time, I suppose it is in > order to tell lies about statistics. > > First, two matters where you are right: > > 1. The UK had year-round BST from > 1968 to 1971. > > 2. During this period there was a > reduction in road accidents. > > That is called an ASSOCIATION. You > cannot infer CAUSATION. There could > be any number of reasons for this > reduction: > > a) The introduction of drink-drive > legislation in 1967. > > b) The increase in seat-belt usage, > not yet compulsory but more and > more cars were being fitted with > seat belts. > > c) Perhaps the winters were unusually > mild. > > Those are the kinds of questions you > should immediately ask before making > ANY inference. > > Next, you absolutely MUST investigate > what happened AFTER we abandoned using > year-round BST. Did road accidents > increase again? > > WHY HAVEN'T YOU TOLD US? > > Answer: either because you didn't look > or because you didn't like what you > found when you did look. > > OK, well look at the attached file: > > RoadDeathsGB.jpg > > You will see that following a post-war > low around 1950, road deaths increased > relentlessly until a peak in 1965. > > They then started a steady decline > which continued throughout the period > 1968 to 1971 and you won't notice so > much as a kink in the line from 1965 > to 1971. The slope in the three years > up to 1968 exactly matches the slope > from 1968 to 1971. BUT: > > AS SOON AS WE STOPPED TELLING LIES > ABOUT THE TIME IN WINTER, THE RATE > OF FALL SIGNIFICANTLY ACCELERATED. > > Using YOUR logic, we can see that during > the winters that we were telling lies, > we actually HELD BACK the reduction in > road deaths. > > YOUR proposal was killing people! > > Funny how the time liars don't tell > us that! > > Now this is still flawed inference > and I wouldn't make this assertion > myself. > > The truth is that it is very hard to > design an experiment to verify one > way or another. You cannot re-run > the same years with different time > rules. > > There is, though, a much better way > to investigate. What you need to > do is to identify two U.S. towns in > the same state that are not far > apart but which are in different > time zones. They should be about > the same size and have the same > mix of population and industry. > > The more easterly of the two towns > will have have its clocks one hour > ahead of the more westerly. By > your hypothesis, there should be > significantly fewer road accidents > in the more easterly of the two. > > OK, go find some research data. > > We can't do this in the U.K. but > you can look at Dover and Calais. > > These towns are fairly close > and the clocks in Calais are > one hour ahead of the clocks > in Dover. > > By your hypothesis, there should be > significantly fewer road accidents > in Calais. > > OK, go find some research data. > > Very best wishes > > Frank > <RoadDeathsGB.jpg> --------------------------------------------------- https://lists.uni-koeln.de/mailman/listinfo/sundial
