This just in: B5 S4 makes top 5 most stolen car!
#1
AudiWorld Super User
Thread Starter
Join Date: Sep 2000
Posts: 9,700
Likes: 0
Received 0 Likes
on
0 Posts
This just in: B5 S4 makes top 5 most stolen car!
not that this should surprise anyone here.<ul><li><a href="http://www.usatoday.com/money/autos/2005-07-19-most-stolen_x.htm">click for article</a></li></ul>
#5
This is a better ranking (makes more sense)...
<ul><li><a href="http://www.statefarm.com/insuranc/auto/nicbcars/grndthft.htm">http://www.statefarm.com/insuranc/auto/nicbcars/grndthft.htm</a></li></ul>
#7
You're right that the sample size is probably pretty small for some of those model/year combinations
and they don't provide theft rates nor indications of how variable the estimates are (e.g. confidence intervals). But I think using a proportion is correct: you can't use just raw number of thefts because it's confounded with car popularity: the most common cars will account for the majority of cars stolen.
Rory
Rory
Trending Topics
#9
AudiWorld Expert
Well, I can see it both ways....
What if EVERY M Roadster was stolen. It probably still wouldn't have the #'s to be as high up as the other cars on that list just because sooo many more of those cars are out there. Wouldn't it seem appropriate that if every M Roadster was stolen, it should make the list? From that perspective, I can see that rankings based on % makes sense. Maybe in combination with %, there should be some minimum #. Or list it both ways. Or at least state the #'s and let us see the info and understand it fully. Anyway...
#10
Ya I agree with your logic. The problem I think is with the article itself...
it says "most stolen" which is incorrect based on the numbers. It should be "most popular" car to steal or something else. Saying it's the "most stolen" misleads the reader when interpreting the numbers.