Originally Posted by
mr_bill
First, there were advisory signs before the "just paint" went down on Lakeview. So, it was advertised and folks on bikes apparently said nope. There was a lot of Waze traffic on that road. With the "just paint" folks on bikes said yep. (For all we know, Waze just gets confused by these "advisory bike lane' streets.) In any event, residents are happy because cut-through traffic has reduced. And people on bikes are happy because it's a safe route.
On Scott (/Irving, it changes name), it used to be a cut-through to avoid most of the Harvard Square chaos. However, now it remains a route if you are riding from Winter Hill (yes, the Winter Hill Gang Winter Hill) to the Charles.
In the meantime, Oxford Street also got the big green sharrows treatment, along with even more recently Quincy Street / Bow Street/ DeWolfe Street changes, so I switched to those streets for my radiation therapy.
As far as your null hypothesis (these changes make no difference), suggest a
double blind study to confirm that these changes are good or not. Oh, wait.
-mr. bill
My null hypothesis?! I should probably have gotten the Nobel for coming up with that being the default test for any statistical statement. I don't think you really can get beyond "well, it isn't doing any harm" using statistical reasoning when there's zero casualties both before and after treatment, which is why I'm suggesting that if what it's doing is encouraging cyclists to choose that zero injury rate route because they feel safer and the injury rate remains zero, that's likely virtue enough.
I guess what I'm unclear on is if you're saying that the advisory lanes have some advantage over big green sharrows. I'm asking
you this because while I have plenty of experience with sharrows in multiple contexts, I have none with these kind of lanes. I think it probably needs to go to anecdotal discussion because, as you point out, there's really no way to have control groups of streets where the conditions of any given road are matched (location matters). Double blinding might actually be the least of the methodological problems, btw. I'm suggesting that if there's a good Hawthorne (people are riding/driving more carefully because it feels like an experiment) or placebo effect (safety is unchanged, but sense of safety is increased), those are probably reasons to do it in themselves.