 |
|
|
[Subscribe to Daily Digest] |
[Profile/Gallery]
, Wed, 14 Jun 2006 04:00:16
Members do not see ads below this line. - Help Keep This Site Online - Signup
you really don't know their research methodology. In a past life I spent 12 years doing market research, 4 of it automotive, and I question how a lot of the data is collected. (I question most survey data in general).
Presumably the JDP surveys were different (although questions were probably similar) as they were asking based on different time periods and the age of the car. JDP would have kept the data from each survey separate from the others so it remained valid. Nothing wrong there.
What they also do (since you got multiple surveys over 2 years) is send surveys to people who respond to surveys. This is to increase their response rate, and there's nothing wrong with that except that it may introduce some bias into the results. How? Well, are people who respond to surveys different than those who don't? There are some reasons to think they may be.
Do responders tend to be more sensitive to stuff going on with their cars than those who do not respond? If so, that may result in things that arent' really "problems" being over-reported.
Are the people who respond more likely to be car enthusiasts or consumer advocacy types? This is a real possibility and is stuff JDP probably has a handle on but isn't made public. But it can weight the results somewhat.
As a car ages and more stuff crops up, are people who are pissed off--perhaps due to a bad dealer experience--more likely to respond than people who are happy? And in so doing are they overy critical or negative? This is not easily answered and introduces some bias into the results that can have a negative impact if the sample size is too small. JDP should control for this, but you can't tell if they do.
At the end of the day most of this probably doesn't matter much, as long as the results aren't seen as absolute facts. The survey data collected by JDP, Condemner Reports and others can at best be generally representative of all cars of a make/model, but it is still the individual car that is the issue. The best thing is to view the data these companies provide as "directional," not stamped in steel. And I don't think it's smart to base a buying decison on a bunch of surveys unless every one of them shows a car to be troublesome.
Lies, damn lies and statistics. Or as Jimmy Buffett would put it, fictional facts and factual fictions.
posted by 24.52.16...
, Tue, 13 Jun 2006 05:32:12
, Tue, 13 Jun 2006 14:20:01
, Tue, 13 Jun 2006 09:40:29, Wed, 14 Jun 2006 04:00:16
No Site Registration is Required to Post - Site Membership is optional (Member Features List), but helps to keep the site online
for all Saabers. If the site helps you, please consider helping the site by becoming a member.
 |
 |
 |
 |
 |
