Quote:
Originally Posted by IK6SPEED
It’s not a random survey of BMW S55 owners so no matter the n= count, it will not be scientific.
You cannot accurately predict margin of error [B]to the real world[B] in a non-random sample survey. You can only predict margin of error to the universe of owners that opted in on Bimmerpost.
|
Might be right - I'm not a statistician, just an enthusiast.
Either way I think we can agree that the survey is of little value in determining the true rate of failure - which is exactly the information someone making the choice of keeping/selling or buying into the M3/4 world would want to know in order to make their decision.
So these results should be looked at with a high degree of skepticism. Still they're likely to be of interest to the community here, especially if there is an obvious trend towards a certain model year, setup, or WHP number. We'll need a good number of responses to get at that information though.