Politi-Score: 2011 First Quarter Preview

Although I've not completed the total counts of the rulings, I'd say at this point I have over 98% of them, enough to publish some preliminary charts.  

The total PolitiFact rulings from January 1 through March 31, 2011, will come to about 480, unless there's some over-production today (March 31) it might get to 500.  About 110 of the rulings came from PolitiFact National.  The preliminary breakdown by ruling is as follows, with the dashed line indicating all rulings with data labels to make it more comparable to the blue "Democrat" ruling line and the red "Republican" ruling line.  I don't expect it will change much from this come April 1. 



The Politi-Score is running 54 for the Democrats and 42 for the Republicans overall for office-holders only (to coincide with Eric Ostermeier's study, although both these charts reflect the total of rulings of PolitiFact National plus the affiliated states).  It appears from the chart above that there was a very wide divergence in the "Mostly True" category with Democrats having twice the percentage of rulings compared to the Republicans.

Another comparison many are interested in is how the averages look in relation to the 2010 Campaign cycle.  Indeed, the truth gets a beating when it's election time:

Politi-Score: Elections 2010 Final Observations

There are some who claim that my Politi-Score is far too generalizing, and there’s some truth to that, particularly from an individual viewpoint. Bryan White has recently written that unfortunately, PolitiFact (PF) does not include a “Disclaimer: These ratings should not be taken as a reliable guide to the subject's truthfulness.” Perhaps I should be including that in my Politi-Score.  As Bryan stated:

Many are tempted to look at the collected data and draw ill-founded conclusions, perhaps like "Kay Bailey Hutchison is just as likely to lie as to tell the truth!"

If this is true, then I would point out that Obama  received 268 PF rulings as of November 1, the largest number of all, and my Politi-Score says the same about him—he averages just over Half True, which means he, like Kay Bailey Hutchinson, is just as likely to lie as to tell the truth. It seems the higher the number of rulings for an individual, the more the Truth-o-meter as an average (Politi-Score) becomes a generalizing, middling, gray, almost equivocating measure, and basically doesn’t tell you much.

On the other hand, I think an example supporting “a reliable guide to” individual truthfulness might be his favorite anecdotal example of PolitiFact’s bias, Michele Bachmann. While I have found that PolitiFact has eased up on the "snark" factor he has so lamented since late 2009, she still is the biggest liar with pants on fire (has the lowest Politi-Score) among those individuals with a large number of rulings, whether by selection bias or not. Often her statements are re-confirmed as false or misrepresentations by more objective sources such as Factcheck.org. Now I wonder, after looking at the infant state-by-state ruling compilations, that if there was a Republican or Independent-leaning “PolitiFact Minnesota” Bachmann might get a fairer shake. In the meantime, yes, she looks to be a “chronic liar”….although it doesn’t seem to bother her constituency, since she keeps getting re-elected.

Politi-Score: Elections 2010 Part Four

Data Analysis by News Source (State)

When compiling the data for this Politi-Score analysis through PolitiFact, I found that the “national” rulings page did not include those of the states. Therefore I had to go through each of the states’ rulings as well (which took a while!). I decided to include state affiliation in the data compilation because I wondered what the “political leanings” (and its affect on the rulings) of these other newspapers were as compared to the St. Petersburg Times, which is often accused of liberal bias by its detractors on the Facebook page.

To get a feel on their political leanings, I looked at who they recently endorsed politically, particularly for president in the last two election cycles. The following table shows the results: some blue (Democrat), one red (Republican), and some mixed (purple).

Location	Newpaper(s)	Endorsements
National	St. Petersburg Times	Obama (08); Kerry (04)
Florida	St. Petersburg Times/Miami Herald	Obama (08); Kerry (04); Miami Herald endorsed Marco Rubio (R) for Senate & Sink (D) for Gov. (10)
Georgia	Atlanta Journal-Constitution	Obama (08); Kerry (04)
Ohio	Cleveland Plain Dealer	Obama (08); no one endorsed 2004; Kasich (R ) for Gov. (10)
Oregon	The Oregonian	Obama (08); Kerry (04)
Rhode Island	Providence Journal	Obama (08); Bush (04)
Texas	Austin American-Statesman	Obama (08); Bush (04)
Virginia	Richmond Times Dispatch	McCain (08); Bush (04)
Wisconsin	Milwaukee Journal-Sentinel	Obama (08); no one endorsed 2004; Walker (R ) for gov. and Feingold (D) for Senate (10)

Politi-Score: Elections 2010 Part Three

Data Analysis excluding Organizations, Blogs and Chain e-Mails

In this analysis, the lower overall ratings of organizations, bloggers and chain e-mails were removed from the Politi-Score averages to see what effect it would have between Republicans and Democrats. When PolitiFact made a Facebook request early in the election season for more liberal chain e-mails from its fan base, I thought perhaps there were a lot more right-wing e-mails, and that this might weigh the averages down for the Republicans thereby favoring the Democrats , since the blogs, etc., have already been found to be, on average, less truthful than the candidates themselves. But as noted in the previous analysis, the patterns were comparable, and although there were more right or Republican rulings, they weren’t enough in number, not too unfairly balanced with the Democrats (43 Republican versus 34 Democrat rulings), and their pants weren’t on fire enough, so to speak, to move the Politi-Score up substantially for the Republicans. In fact, the effect was almost nil, overall raising the "truthiness" scores for each party roughly five percent, using the wide measure. It was a miniscule under-one-percent using my previous grading scale.

Note:  There were twenty additional rulings which could not be identified with either party.  They were included in the total average (#Rulings and Overall columns) but are not broken down.

So far, similar pattern. Looking at the news sources (PolitiFact states) might yield more results worthy of note. That’s next. 

Politi-Score: Elections 2010 Part Two

Data Analysis of Organizations, Bloggers and Chain e-Mails

Imagine this conversation in a Bill Adair PolitiFact Staff meeting circa late August this year: “Now this is how I want our rulings to work on anyone other than the candidates for the upcoming campaign season….they should be about 20 percent of all rulings..and try to keep it equal between the Republicans and Democrats if you can. Overall, they should probably be less truthful than the candidates, so don’t be surprised if you have a lot of Pants on Fire……hmmm…. maybe 3/8 False and Pants on Fire, about half in the middle, and the rest True or Mostly True?” Well, I can imagine him making every part of that statement (to get to an ultimate "Barely True") except the last one, after the thoughtful “hmmm.” 

At least that’s how it worked out, almost with an uncanny precision: Of the 414 PolitiFact rulings posted between September 1 and November 1, 80 of them, about 20%, were by organizations (ranging from Karl Rove’s Crossroads GPS to “Blue Oregon”), blogs (one simply called “Bloggers”), and chain e-mails (one also simply called “chain emails”). A little over half were from Republican groups. Ten percent of the rulings were equally divided between True or Mostly True, 52.5% were equally divided Half True or Barely True, and the rest (3/8 or 37.5%) equally divided between False and Pants on Fire.

Politi-Score: Elections 2010 Part One

Data Averages Overall

PolitiFact (PF) recently published its own evaluation of the recent mid-term election campaign, calling it, in Truth-o-Meter terms, a "Barely True."  So, was it?

Well, a bit of time was spent was going through the ratings one by one from September 1 through November 1 (the day before the election) to see if I could validate Bill Adair's calculations.  My numbers came out a bit different than his.  Since he published his evaluation on October 27, I probably counted additional rulings that slightly changed the percentages that he calculated.  I counted 414 rulings in total.  This was a bit different from the way I had initially surveyed, which was by selecting only those individuals who had over five (5) rulings by going through the list as shown under PF's "People" tab.  This time it was a total cross section of two months of all rulings, including organizations, bloggers and chain e-mails.

The result was an agreement with PolitiFact on "Barely True" in total.  But further dissection of the numbers to divide them between Democrats and Republicans revealed who was, as judged by PolitiFact, more than Barely True.