Well, at least he didn’t call it “skewed.”
But let’s look at the “bogus” claim for a minute, lest we concede it’s a business-as-usual complaint by a Republican pol against a Democratic polling firm.
The question involved was a straightforward job approval favorability or unfavorability inquiry. No questionable wording was involved, and it appears McCain’s job approval came up first in a list of Arizona pols, so question order shouldn’t be an issue.
Now most often poll results are questioned over the composition of the sample: too many of these, not enough of those. But McCain’s job approval ratios were pretty even across party ID lines: 29/53 among Dems, 35/55 among GOPers, 25/55 among indies. A shift in sample composition wouldn’t have made much difference. Could the sample have perhaps been too small? Well, the poll came with an MoE of 3.3% to reflect the sample size, and again, a maximum shift in McCain’s favor would have left him underwater in all three types of voters.
What’s McCain’s counter-data?
I can sense the people of my state. When I travel around, which I do constantly, they like me, and I am very grateful.
Oh, okay. That’s convincing.