"If they have been weighting by party, and they suddenly changed their weighting, I'd say it's a pretty questionable practice. The last four Fox surveys all varied slightly in terms of party ID, so I'm not sure that's what they've done. That said, if the other surveys this week show no similar "shift" in party, then it's probably reasonable to discount this result."
The weighting was dramatically different from the last poll - and yet their news story compared the two as if they were identical samples. Nate: your take?