Upgrade your account to a Community Supporter account and remove most of the site ads.

Reply to thread

Message: <blockquote data-quote="Mark CMG" data-source="post: 5631767" data-attributes="member: 10479">Actually, it *isn't* a caution that negates the collective opinions from being accurately representational when the sample size is large enough.  But you've chosen to only quote part of what I said.  Since you ignored a portion of my post where I discussed the sample size being large enough, you've managed to only be correct in regard to what you have posted, not in regard to what I posted in whole.  The larger the sample size, the less likely self-selection is a factor in whether the data is qualitatively representational.  Furthermore, as to whether it is quantitatively accurate, the larger the sample size, the less likely it is inaccurate as well though any individual opinion is still more difficult to assert as being representational of the same percentage of the whole.  In any event, one has to determine if the sample size is large enough to feel the data can be accurate no matter whether you are dealing with a self-selected or a randomly selected sampling of the whole.The existence of bias can not be predicted as more or less in evidence in self-selected groups then in randomly selected groups.  And in point of fact, there is no way to eliminate bias in any type of group.  You also cannot say definitively that the biases  of any group, self-selected or otherwise, aren't possibly an accurate representation of the biases of the whole.  No matter the sample size or the manner of selection, it is paramount to collect large amounts of data and use polling questions that are fashioned to lessen the impact of biases.  To offhandedly toss out self-selection in a manner that overstates its importance whenever a discussion of how representional message board opinions can be is a narrow, incomplete view of the science of statistics.</blockquote>

Verification