Date: Sat, 04 Aug 2007 19:56:36 GMT From: still me <wheeledBobnospamo.com> Subject: Re: 2003 9-3 with 78,000kms - am i making a mistake?
On Sat, 04 Aug 2007 15:24:34 -0400, Konan <kohoonospamo.net> wrote: >bozo wrote: >> A couple of other things to keep in mind - Consumers' Reports is basing its >> statements on comments and observations made and reported by owners, > >Is there a better way to get facts? Yes, in a statistical survey. I'm guessing you didn't take statistics in college. That's not an insult, just that if you never studied statistics, you might not realize how corrupt their methodology is. The first rule in statistics it that to get a truly accurate sampling, your sample has to be truly random. Their's is no where near that. What they do is solicit opinions from their subscribers. There is no attempt to draw an accurate random sample from the population of vehicle owners. In addition, they take opinions, not repair records and call them facts. It just gets worse from there, as they draw averages and ratings from totally invalid samples. From a statistics point of view, what they do is pure bunk. It leads to things like the famous headline "Dewey Wins". Except CR is worse. To add to that, even with good samples, their data is very misleading. As an example, they have a category for major engine problems. Unfortunately, some of the things in that category are very minor fix, and some are major repairs. For example, the Nissan stud problem I mentioned earlier is definitely an issue, but it is fixable with no permanent damage. OTOH, Mitsubishi had a problem with "crank walk" in their engine design for a few years wherein the clutch pushed the crankshaft back and forth until the crankshaft eventually got so loose it walked it's way right through the front of the engine. Big difference in seriousness of the problem. But CR doesn't tell you what kind of problem it is, they just give you a less than full red circle.