If their poll results skew heavily toward one party or the other, when all other factors indicate a much narrower margin, then they should examine their methodology and sampling to find the source of the error, rather than rushing to print with it.

The fact that Newsweek chose to ignore the warning flags instead of revisiting their polling procedures tells me that they are either terminally sloppy or reliably biased. Given the body of their past work, it’s not difficult to conclude the latter choice is the true one.

“Do you consider yourself a Republican, Democrat, or Independent?”

I’m no expert on polling techniques, so I’ll listen to anyone who can straighten me out on this, but doesn’t the question suggest that the party affiliation was determined after the person chose to participate in the poll, rather than before? Also: in light of the recent election results, wouldn’t one expect to see more people now self-identifying as Democrats? That would go along with the fact that so many Democratic candidates were elected.