Reply to thread

Message

<blockquote data-quote="Umbran" data-source="post: 5631694" data-attributes="member: 177">Actually, it *is* a caution that negates the collective opinions from being accurately representational.&nbsp; That's the science of statistics.&nbsp; The self-selected group will generally be quantitatively and qualitatively different from the general population.&nbsp; To say that a group represents the whole is more than just &quot;all opinions that appear in the whole appear in the group&quot;.&nbsp; To be properly representational, the opinions of the whole appear in the group with the same frequency, with the same vehemence, and the group should not have opinions that don't appear in the general population. If you poll a self-selected group, and you poll the general population, the poll results will usually be different.&nbsp; That means the one does not represent the other.&nbsp; You can even often poll a self-selected group on things that seem entirely unrelated to the selection (for example, asking self-described &quot;fiber artists&quot; about politics) and still not have it match the general population.&nbsp; The biases can run deep and strange.</blockquote>

[QUOTE="Umbran, post: 5631694, member: 177"] Actually, it *is* a caution that negates the collective opinions from being accurately representational. That's the science of statistics. The self-selected group will generally be quantitatively and qualitatively different from the general population. To say that a group represents the whole is more than just "all opinions that appear in the whole appear in the group". To be properly representational, the opinions of the whole appear in the group with the same frequency, with the same vehemence, and the group should not have opinions that don't appear in the general population. If you poll a self-selected group, and you poll the general population, the poll results will usually be different. That means the one does not represent the other. You can even often poll a self-selected group on things that seem entirely unrelated to the selection (for example, asking self-described "fiber artists" about politics) and still not have it match the general population. The biases can run deep and strange. [/QUOTE]

Verification