Quote:
Originally Posted by regular guy
I think the best thing to do is to trust people if they say they are posting for another who isn't on the forums.
|
It depends on what you want to achieve....if its just a big number with not much value, then anyone who has heard of a failure should vote and lets see how high we can get it.
OTOH if want a result that you can use to try to predict the % failure rate then only the owners of a car that has had an engine failure should have voted.
Analogy: You want to get an idea how many of the population smoke...so you ask a random sample of people if they smoke and use the results to extrapolate to the entire population. You don't specifically invite known smokers into the sample. You don't ask the sample if they know of anyone who smokes and you definitely don't invite a tobacconists into the sample and ask him how many people he knows who smoke....unless of course you want skewed results for some particular reason.