Quote:
Originally Posted by jhautz
...No... but just about all. And it shouldnt matter. as long as we can get a statitically significant sample size we should be able to estimate the good/bad percentages within a reasonable number. The ESC showed up to use randomly so the failure rate should represent a random sampleing...
|
The poll is statistically accurate as long as the same % of people with failures post as the same % of people with success. I still think there are more successes than are being reflected here. How many more? Who knows? The only true gauge of that would be to tabulate the returns at Castle.
IMO, this poll is just
roughly indicative of the failure rate.