
Originally Posted by
Ixion
I found the reason for the anomaly , after discussion with an ACC dude.
The figures used in the capacity calculations are a (non random) subset.
They are only serious crashes! No wonder they are higher.
The selection bias is because, for the capacity calculations , they linked to the Police CAS system , which is the only place they can get the capcity figures.
Then tried to match up the ACC crash record with the police record. Only a quarter matched.
Since Police attend serious crashes but (usually) not minor ones, the police recorded crashes will be those where there are serious injuries .
For teh cars they just took the total number of crashes.
Bookmarks