Here's a real shock -
USA Today presents a table demonstrating that when polls include disproportionate numbers of people from one party or the other, the results of the poll skew toward the overrepresented party. With all due respect to
Gallup's rationalizations for releasing polls which overrepresent self-identified Republicans, I think critics do understand the science (or lack thereof) behind this type of polling.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.