Originally Posted by
wphamilton
This seems odd: daytime running lights (DRL) on passenger cars appear to
increase the number of collisions according to this study of US crashes.
The Effectiveness of Daytime Running Lights For Passenger Vehicles There was also found an adverse effect on car/motorcycle crashes when the car had DRL. Trucks and vans with DRL saw a reduction in crashes.
Certainly passenger cars are not bicycles, but the result does imply that what seems obvious on the face of it (that DRL in traffic increase visibility therefore reduce crashes) isn't necessarily sufficient for the conclusion.
Read it again.
Two-Passenger-Vehicle Crashes Excluding Rear-End Crashes (Target Two-PV Crashes)
- ...
- DRLs significantly reduced the LTVs’ involvement in daytime target Two-PV crashes by 5.7 percent at the 0.05 level.
- The remaining results were not statistically significant at the 0.05 level.
- ...
Single-Passenger-Vehicle-to-Pedestrian/Pedalcyclist Crashes (Single-PV-to-PED/CYC)
- ...
- None of the results were statistically significant at the 0.05 level.
Single-Passenger-Vehicle-to-Motorcycle Crashes (Single-PV-to-Motorcycle)
- ...
- All the results were not statistically significant.
- ...
A *correct* summary of the report is:
- There is little evidence from analyzing the FARS and "State data" that DRLs are effective.
- But there is *also* little evidence from analyzing the FARS and "State data" that DRLs are ineffective.
- And there is *also* little evidence from analyzing the FARS and the "State data" that DRLs are harmful.
Opinion follows:
FARS and "State data" are *noisy* data sets. DRLs would have to be nearly magical to see their effect in such *noisy* data sets.
Are DRLs good, bad, or neither? From looking at these *noisy* data sets, who the hell knows?
It may however be safe to conclude from this study that DRLs are not magical.
Recommendation for future work.
Dramatically improve the quality of FARS and "State data"
- OR -
Use another methodology for examining effectiveness of DRLs.
-mr. bill