"[They] run a regression that states more miles driven the less likely an airbag is to deploy. They state this is unexpected but don’t explain what was wrong wi

"[They] run a regression that states more miles driven the less likely an airbag is to deploy. They state this is unexpected but don’t explain what was wrong with their data to cause it" This is covered in the Discussion section, but not with much clarity. I think it makes more sense if you invert it: cars that have been in a crash accumulate fewer miles than cars that have not been in a crash This effect seems obvious to me, since a car can't be driven while it is being repaired (and will never be driven again if it was totaled).