On Sun, 19 Mar 2017 11:53:38 -0700, sms
wrote:
I know how some people dislike any statements that are based on actual
facts, but the data are pretty clear.

See: https://books.google.com.au/books?id=LvthAgAAQBAJ&pg=PA168

It was interesting that in the Odense study, conducted by Reelight,
^^^^^^^^^^^^^^^^^^^^^
Odense Cycle City and the University of Aalborg, accident rates went
down by 32% with the use of daytime lights, but a cyclist's "sense of
security" went up by 85%!

It's funny how often research just happens to find what the person
paying for the research would like it to find (unless you're a
Republican politician paying for climate change research).

That said, I wouldn't be surprised if daytime flashing lights did have a
net positive effect on rider safety. Seems to me the problems with
those are when they are a rider's only light at night.

My car has daytime running lights set to be on by default. Motorcycles
have headlights on when riding during the day or night by law here. The
research on DRLs is not especially consistent but seems net positive,
sometimes barely so and sometimes significantly so:

http://www.dot.state.mn.us/research/...11/TRS1009.pdf

"The most recent large-scale study on this topic conducted in the United
States is a 2008 NHTSA study that found that DRLs had no statistically
significant effects on the types of crashes studied, except for a 5.7
percent reduction in the involvement of light trucks/vans in two-vehicle
crashes. A 2004 NHTSA study that used different analysis methodology
found that DRLs reduced opposite-direction fatal crashes by 5 percent
and opposite-direction/angle non-fatal crashes by 5 percent. That study
also found a 12 percent reduction in crashes involving pedestrians and
bicyclists, and a 23 percent reduction in opposite-direction crashes
involving motorcyclists."

Of at least some interest to us, there may be some evidence that DRLs on
cars might reduce your chances of getting killed in a collision with
one. None of the studies cited seem to have tried DRLs on bicycles.

But the article also points out that methodology has an effect on the
existence, direction and magnitude of any particular outcome. A
seemingly large result with one methodology might not be statistically
significant with another.