Yeah (alliteration alert), the folks at Fox & Friends are smarmy and smug. They are also, for reasons mysterious, popular. This morning, as I scanned the various news channels to catch what was breaking, the Foxers mentioned their own poll that showed Obama leading Romney both nationally and in three key swing states.
That the Fox friendsters would report a poll that looks good for Obama -- surprising.
That Fox friendsters would then spin a poll to make Romney look good -- not so surprising.
These Fox News polls are sound methodologically. What's interesting is how the smarmeisters on Fox & Friends, after noting Obama's lead, then pointed out that among respondents who are "extremely interested" in the prez election, Romney scoots ahead of Obama.
Here's the problem. Take their Florida poll, for example (detailed results here). About 6 out of 10 in that state report being "extremely interested" in the campaign. The total survey was of 829 respondents. Fifty-nine percent of that = 489 people. The margin of error? Call it about 4.5 percent, meaning you'd have to be more than 9 percentage points ahead of the person for it to not be a statistical tie.
Romney's sudden vast lead among these people? Well well well within the margin of error.
The friensters at Fox of course glossed over this little methodological quirk, I suppose to make the Romney results more palatable to their discerning audience.
Don't get me wrong: I'm bitching here about the misuse of poll data, not who is ahead or behind in the poll. It's the poll abuse I dislike, especially when trying to spin the results a certain way. Yes, we all torture data sometimes, but in this case it's waterboarding to the extreme.