Quote:
|
Originally Posted by Wisha Haddan H3
This does not discredit the study. There is good reason to include nighttime crash statics in an analysis of DRLs. For example:
1. Many DRLs are controlled by a light sensor, which automatically turns on full wattage and activates marker lights during dusk/dawn/night. With nothing to forget, the driver's risk could decrease.
|
What percentage of cars do you see driving around at night with no lights on? It is not zero, but it is pretty close to it.
Quote:
4. Including nighttime statistics also eliminates the variables of winter/summer daylight hours and the need to define "dusk" and "dawn".
|
It doesn't eliminate variables. It ADDS a control variable. The statistics shouldn't change one bit with or without DRLs at night. If it does, you know there is noise in the data, and you use that to build a confidence band.
Michael