Originally Posted By: Kestas
Though JD Powers and CR are less than perfect, and we seem to make a sport out of bashing their results, I think these surveys are the best we have. I can't think of anything better that we can turn to.
A notice at the bottom of every page in the press release specifically says that statistical significance is not indicated in the ratings. This means that the ranking is largely meaningless. You can be vaguely sure that the numbers near the very top of the chart are better than the ones near the very bottom, but that's only based on intuition. Bad data is not better than no data.
However, let's give them the benefit of the doubt.
Assuming the data is valid, the value of the survey depends on how you take it. You have to keep in mind that what they don't say is just as important as what they do say.
Let's take this study as an example. What it says is that, in a given year, you'll see a lot more broken VWs than broken Buicks. If I were thinking of setting up a repair shop, this data would help me decide which cars to focus on. If I were thinking of opening a dealership, I might want to take it into consideration as well.
On the other hand, the study doesn't say anything about how the cars do with mileage, which parts break, and how expensive and time consuming the repairs are. It only considers the age of the vehicle and doesn't discriminate too much on the nature of the problem. So, if I were buying a car, I would not find the study particularly informative since those things are very important to me.