Zebra Factcheck is the PolitiFact *Unskewed*

(OR "Fact-checking you do as a Republican to make yourself feel better"....)

If there's one thing to be learned from the elections, it's that creating one's own realities doesn't work. From Salon.com:

Conservatives will hear that Hurricane Sandy and Chris Christie delivered the election to Obama, that Mitt Romney wasn’t a true conservative, and that mainstream media bias blinded Americans to the truth. A few voices may acknowledge that the state of polling science is actually pretty good, and the demographics of the U.S. have fundamentally changed, but don’t hold your breath for a widespread reappraisal of the dangers of listening only to what you want to hear. The right-wing echo chamber is built of stronger stuff than the left’s.

And along with "mainstream media bias," fact-checkers must be 100% liberally biased as well! And Zebra Factcheck will unblind you to "the truth"! But I digress.

Current TV's Cenk Ugyur reported how much 26 polls were off in predicting this past election, listed in terms of accuracy: it almost came as no surprise to me that one of those near the bottom was the conservative Rasmussen--while the top poll was the liberal Public Policy Polling and one from the Daily Kos.

 

There were a number of polling analysis outfits, the most prolific being Nate Silver's FiveThirtyEight, along with Real Clear Politics. But then a website called Unskewed Polls became well known, especially with Republicans, with over 25,000 fans on Facebook, as authored by Dean Chambers who was very highly critical of Nate Silver. It essentially based its election prognostications adjusting for what it believed was over-sampling of Democrats by the other polls (and under-sampling Republican enthusiasm)--because those polls were part of the "main stream media" and that by "showing results unrealistically favorable to Obama that they are leading to a perception among voters of the inevitability of Obama's victory in the election." 



Unskewed Polls electoral map Nov. 4: Romney to win 52% to 47%

And so, the weekend before the election up until election eve, Unskewed Polls was basically predicting a Romney sweep of 342 electoral college votes to 196 for Obama, as shown above  in a screen-snip from its website. Sometime Election Eve Monday Dean Chambers did some quick back-pedaling (strangely, with almost all the polling done) on his electoral map to make it look like he only failed to predict four states for Obama, changing the electoral counts to 275 for Romney and 263 for Obama. Meanwhile, people like Karl Rove and Dick Morris were making similar errors to Dean Chambers on Democratic over-sampling, in predicting a Romney landslide on Fox News. Everyone ended up eating a lot of crow failing to discern the voter suppression backlash and growing minority demographics.

 

So you can see where I am going with PolitiFact Bias (PFB) and its new "project" Zebra Factcheck. I believe the writers at PFB commit the same type of cognitively dissonant blunder, insisting that PolitiFact is "Democratically skewed" like Chambers claims all the polls are. And they think they are unskewering PolitiFact's rulings at their websites, just as Chambers says he does.

A Case for PF Wisconsin Right Wing Bias

While some try as best they can to prove that PolitiFact National has a left wing bias, there are a number of good reasons why it most likely doesn’t exist at national level.

1. The selection bias allegation is assumed to be only a party-related bias, contrary to PolitiFact’s repeated claim that is a newsworthy bias, and that ALL PolitiFact National writers/researchers/editors are liberal (and have dispensed with journalistic training in objectivity), for which one of its most vocal critics has provided NO evidence except for his own opinions of an anecdotal review of its rulings,  tainted with his own admitted conservative bias.

2. That other fact checkers, such as FactCheck.Org and the Washington Post FactChecker, check many of the same claims and (about 99% of the time) make similar findings, so therefore, it would have to be proven they are liberally biased as well in reaching the same conclusions.

3. There appears to be no obvious agenda cited by the critics except the sweeping generalization of selecting only statements favorable to Democrats and not favorable to Republicans.

IN WISCONSIN, however, using the above three points, there’s a very compelling situation which makes a good case for Republican (or right-wing) selection bias: .

1. A limited number of writers (they could be biased—but again, there’s no evidence they are, but it seems it would be less possible where there are more than just a few writers, as at PolitiFact (PF) National).  As to the question "who checks the fact-checkers" PolitiFact Wisconsin Senior Editor Greg Borowski could only say "all of our sources are listed and linked.  So check away."  But is having sources listed sufficient?

2. No one else fact-checking same claims because they are local to Wisconsin. .

3. An obvious agenda (the recall election) reflected by a Pants on Fire on the Democrat contender for governor on the eve of the election, with no claims checked on the Republican, and a few days before, particularly June 1, claims reported rated much more positively for Republicans.

4. And I’ve already noted that Republican-leaning news organizations in the PolitiFact domain tend to have truth indexes which favor the Republicans more. The Milwaukee Journal Sentinel, which sponsors PolitiFact Wisconsin, is a somewhat mixed publication in its endorsements: it endorsed Walker, but also endorsed Obama in 2008 and did not endorse either Bush or Kerry in 2004.

5. Individual Truth Index trends: Prior to the 3 month period before the June 5 Recall, Scott Walker’s Truth Index was a negative 43, while in that three month period it improved modestly to a negative 21. Tom Barrett’s Truth Index up until that three month period was a negative 9. In the three months prior to the Recall, it plummeted to a negative 62. Mind you, this was for only eight fact-checks.