Reggie

But CR takes out the bias (one way or the other)

BTW my survey makes the A4 B5 look good (sic)

April

data from a number of repair facilities. Year of car, engine, type of problem, etc. Rather than the owner of the car giving their interpretation of the problem (and let's face it, most people are clueless about cars), the data comes from the facility that diagnosed the car.

In the case of CR, if you end up with a small sample set and by some chance a few of them end up having large issues (luck of the draw), then the survey can be easily skewed. IMHO, you need a large sample set for something like this to be accurate.

I would love to see something like the German system used in North America.

Reggie

I would not trust that data.

April

I don't know how many times an owner of a car has told me something they think is technical, only to have it not jibe or be entirely impossible for it to be correct. ie: stereo affecting functionality of the a/c this side of a massive electrical problem.....

Usually the owner of a car is using the repair order as the starting point for their description of what is wrong with a car anyways, so by eliminating the owner and simply jumping to the repair order, you eliminate one possible source of mis-interpretation. That is unless you are surveying the ability of the repair order to describe the complete condition of the vehicle and then the (usually non-technical/mechanical) owner's ability to convey the gist completely;-)

April

<ul><li><a href="http://www.thetruthaboutcars.com/?p=2043">takes a while for the whole thing to catch up - good or bad</a></li></ul>