Redactors Does An Analysis Of Politifacts.com

[Moonbeam]

Just stating an opinion and watching people getting their panties in a bunch in their rush to refute 'nothing of substance'. Perhaps your question is best answered by those who salivate when they hear the bell ring.

But your opinion was the bell ringing.

 

[Doc Savage Fan]

Far too small a sample size and a focus on a rare event (the 'pants on fire' rating). This is a bad/invalid use of statistics by "conservativefactcheck.com" from which to draw conclusions about Politifact's leanings. This is likely due to their own bias, which is pretty funny considering the accusation, and also funny considering the statement that 'people like to hear what they want to hear'.

Not to mention that it holds Politifact finding that conservatives say more outrageous and false things as prima facie evidence that they are biased. That's circular reasoning. The chart looks pretty right to me.

For fun, go look at the statements rated 'pants on fire' that Romney or Michele Bachmann said. They are all absolutely ridiculous.

http://www.politifact.com/personalities/michele-bachmann/statements/byruling/pants-fire/

http://www.politifact.com/personalities/mitt-romney/statements/byruling/pants-fire/

That's the complete list of ALL "Pants on fire" ratings ever recorded by Politifact as of 2012. If that chart screams objectivity, integrity and credibility to you...then so be it, as I would figure as much.

 

[Moonbeam]

Ergo my mentioning that some Dems are also complicit in what the Repub majority wants to do. I guess I didn't stress that enough in my post.

To clarify what my post was about and in response to your retort, just let me say that it's the preponderance of the Repubs in Congress that initiate and pursue their objective of siding with and working for the very wealthy, with the minority of Dems who are complicit going along for the ride.

Check the proposed legislation of each party coming out of Congress when they are in the majority for proof of what I'm opining about.

There have been numerous occasions where the Dems have had to cave to Repub demands in order to get the tiniest of concessions out of them. It's part and parcel of the set in stone trend where the Dems have always had to give a lot to get a little. It's a true and accurate indicator of how the very wealthy have so much more sway over Congress than does the middle class and the poor. It's also a perfect gauge of how utterly corrupt many of our Congress critters from both parties have become.

But let me reiterate a significant point of contention once more: It's the overwhelming majority of Repubs in Congress that are passing legislation that exclusively favor the very wealthy with a few yet significant number of Dems siding with them to get their legislation pushed through into law, with, of course, a return of the favor for those Dems to get their own pet projects pushed through.

The Dems, for the most part, aren't introducing legislation that exclusively favor the rich, and I'd hazard to guess with high confidence that when the Dems do it, it's to get something small for the middle class and the poor in exchange for it.

edit - Let me add that there have been numerous occasions where I felt that the Dems in Congress had betrayed their constituents quite badly, and that the sacrifices they made were nowhere equal to the gains they got out of it. In comparison, it's my opinion that almost every piece of legislation the Repubs have passed has been in favor of the very wealthy at the expense of the middle class and the poor.

Why not, if you really really hate yourself, pretend to be superior to everybody rather than just one or the other party. You get to cast an airy-fairy vote.

 

[Doc Savage Fan]

But your opinion was the bell ringing.

I've asked you a few questions lately in various threads and was disappointed you chose not answer. What would you have me say to yours?

 

[Moonbeam]

That's the complete list of ALL "Pants on fire" ratings ever recorded by Politifact as of 2012. If that chart screams objectivity, integrity and credibility to you...then so be it, as I would figure as much.

In your own words, if that chart doesn't scream objectivity, integrity, and credibility to you.... then so be it, as I would figure as much.

It all comes down to the certainty you have in your feelings of how things look.

I don't have an opinion and no idea of the truth, probably why I see slant.

 

[fskimospy]

That's the complete list of ALL "Pants on fire" ratings ever recorded by Politifact as of 2012. If that chart screams objectivity, integrity and credibility to you...then so be it, as I would figure as much.

I don't think you understand what I'm telling you. 'Pants on fire' ratings are quite rare; drawing conclusions from improbable events with very low sample sizes is a big statistical no-no. Since you seem to be drawing conclusions on objectivity, integrity, and credibility from this I should tell you that you are doing so foolishly.

Additionally, nearly half of Mitt Romney's 'pants on fire' claims are about the same few statements that he kept repeating even after they were shown to be outrageously false. Counting them multiple times in an analysis would also probably be a bad idea. Finally, you continue with the idea that just because Politifact has rated more Republican statements to be outrageous in recent years that this somehow undermines their integrity. That is a fundamental analytic error, because it assumes that the correct balance should be 50/50 (or so) without any support for why.

For these reasons and more no researcher would say that was a good idea to try to do what you're doing with the data you have presented. If you insist on doing so then I must conclude that you have interests other than determining if Politifact is objective, credible, and has integrity. Do you?

 

[Moonbeam]

I've asked you a few questions lately in various threads and was disappointed you chose not answer. What would you have me say to yours?

If so it is because I missed them. I don't ignore you at all. I think I would love being your friend and knew you personally. I think you have known pain that would make my soul weep.

 

[Doc Savage Fan]

I don't think you understand what I'm telling you. 'Pants on fire' ratings are quite rare; drawing conclusions from improbable events with very low sample sizes is a big statistical no-no. Since you seem to be drawing conclusions on objectivity, integrity, and credibility from this I should tell you that you are doing so foolishly.

Additionally, nearly half of Mitt Romney's 'pants on fire' claims are about the same few statements that he kept repeating even after they were shown to be outrageously false. Counting them multiple times in an analysis would also probably be a bad idea. Finally, you continue with the idea that just because Politifact has rated more Republican statements to be outrageous in recent years that this somehow undermines their integrity. That is a fundamental analytic error, because it assumes that the correct balance should be 50/50 (or so) without any support for why.

For these reasons and more no researcher would say that was a good idea to try to do what you're doing with the data you have presented. If you insist on doing so then I must conclude that you have interests other than determining if Politifact is objective, credible, and has integrity. Do you?

Do you understand what selection bias is? That chart screams it and your attempt to chalk it up as a mere statistical aberration is patently absurd. Joe Biden has only said 3 outrageous and false things in his career?!? Really? Face it, Politifact views the political world through a liberal lens. Time to deal with it.

 

[michal1980]

Do you understand what selection bias is? That chart screams it and your attempt to chalk it up as a mere statistical aberration is patently absurd. Politifact views the political world through a liberal lens. Deal with it.

he wont.

If it confirms his bias its ok.

If it supports a different pov, then its not worth reading.

 

[fskimospy]

Do you understand what selection bias is? That chart screams it and your attempt to chalk it up as a mere statistical aberration is patently absurd. Politifact views the political world through a liberal lens. Get real.

I understand selection bias much better than you do I'd imagine.

As I mentioned before, the chart only screams it if you have some reason to expect that the number of 'pants on fire' statements should be roughly equal between the two parties. Considering that the party out of power and parties attempting to appeal to radicalized bases, especially in contested primaries have much greater incentives to say outrageous things, I would say that there is zero reason to think that they would be roughly equal.

How did you come to the conclusion that they should be? Please be as specific as possible.

Also, do you think you are looking at Politifact in an objective manner or do you think you're looking to find reasons why what they have done is wrong so you don't have to accept it?

 

[brycejones]

Do you understand what selection bias is? That chart screams it and your attempt to chalk it up as a mere statistical aberration is patently absurd. Politifact views the political world through a liberal lens. Deal with it.

That chart does not scream selection bias since it is narrowly focused on one potential outcome.

To even begin to show the potential of selection bias I would think you need to show that overall they rated more statements from members of one party over the other. Do you have evidence of this?

 

[FerrelGeek]

Believe me, I give it plenty of thought. You're correct in what you said. Doesn't mean I have to be happy about it. Doesn't mean any of us should be happy about it.

Give it some more thought then.

Think about what's involved, and why anyone would do that to themselves. The pool is automatically small and warped.

 

[realibrad]

he wont.

If it confirms his bias its ok.

If it supports a different pov, then its not worth reading.

The vast majority of people do that on this forum. You did that recently with an AP article.

 

[brycejones]

he wont.

If it confirms his bias its ok.

If it supports a different pov, then its not worth reading.

Name a website you find acceptable that performs the same function as politifact or factcheck.

 

[Sheik Yerbouti]

Name a website you find acceptable that performs the same function as politifact or factcheck.

HA!

 

[fskimospy]

That chart screams it and your attempt to chalk it up as a mere statistical aberration is patently absurd. Joe Biden has only said 3 outrageous and false things in his career?!? Really?

The question is not 'has Politifact successfully cataloged all the false and outrageous statements that have been said?', it is 'do Politifact's ratings accurately represent the relative pants-on-fire-ness of the two parties?' I mean seriously, 12 outrageous and false things for Michele Bachmann is probably the product of a particularly busy afternoon, yet she only has 12 entries when she could easily have hundreds.

It is, again, foolish to even attempt to draw meaningful conclusions from such small samples of such improbable events, but to even start you would need to, again, come up with some conclusion as to what the distribution SHOULD be. You appear to think it should be around 50/50, but have failed to provide any evidence for why. I would say that I would be suspicious of any result that WAS around 50/50, considering the dynamics.

 

[trenchfoot]

Why not, if you really really hate yourself, pretend to be superior to everybody rather than just one or the other party. You get to cast an airy-fairy vote.

A point well taken. Thanks for mentioning that. :thumbsup:

 

[Doc Savage Fan]

Also, do you think you are looking at Politifact in an objective manner or do you think you're looking to find reasons why what they have done is wrong so you don't have to accept it?

I like Politifact and use it relatively often. Conservatives have complained about selection bias for years and I gave you evidence that shows that this may indeed be the case. When only 17% of liberals warrant a "pants on fire" rating vs 83% for conservatives (5 to 1 ratio), this infers one of three possibilities: (1) that liberals are 5 times less likely to say ridiculous lies as conservatives, (2) the site is disproportionately fact-checking conservatives and grading them more severely, or (3) both (1) and (2) are true to some degree. You chose. Believe what you want to believe, disregard the rest.

 

[Doc Savage Fan]

The question is not 'has Politifact successfully cataloged all the false and outrageous statements that have been said?', it is 'do Politifact's ratings accurately represent the relative pants-on-fire-ness of the two parties?' I mean seriously, 12 outrageous and false things for Michele Bachmann is probably the product of a particularly busy afternoon, yet she only has 12 entries when she could easily have hundreds.

It is, again, foolish to even attempt to draw meaningful conclusions from such small samples of such improbable events, but to even start you would need to, again, come up with some conclusion as to what the distribution SHOULD be. You appear to think it should be around 50/50, but have failed to provide any evidence for why. I would say that I would be suspicious of any result that WAS around 50/50, considering the dynamics.

Probably should have been much more lopsided. Just a statistical aberration. Got it .

 

[michal1980]

The vast majority of people do that on this forum. You did that recently with an AP article.

but he thinks he's smarter then everyone.

 

[michal1980]

Name a website you find acceptable that performs the same function as politifact or factcheck.

I dont think I criticized politifact findings. But I have a problem with the further reddit analysis.

Because while politifact might be correct in rating stories true or false. The stories it chooses to rate can change the overall mix of ratings.

IE politifact might be choosing more republican stories that will be rated false, while ignoring stories where democrats are false.

 

[Doc Savage Fan]

but he thinks he's smarter then everyone.

He's pretty smart.

 

[Doc Savage Fan]

More proof that Politifact is remarkedly unbiased!

http://mediatrackers.org/florida/20...-times-scores-pants-on-fire-for-partisan-bias

Since 2000, the Times has issued 10 endorsements in elections for U.S. President, U.S. Senate, and Florida Governor. Nine of the 10 endorsements went to Democrats, with the sole exception being the Times’ endorsement of Democrat-leaning Independent Charlie Crist in the 2010 U.S. Senate contest.

The Times’ Democrat bias is perhaps most apparent in its endorsement of Bill McBride in the 2002 Florida Governor election. After picking up substantial Democrat and Independent support in a landslide 11-percentage point victory over Democrat Buddy MacKay in the 1998 gubernatorial election, Republican Jeb Bush earned high grades and bipartisan respect in his first term as governor. As a result, Bush cruised to an even larger 13-percentage point landslide victory in his 2002 reelection bid against Democrat Bill McBride. Despite Bush’s strong record and bipartisan support, the Times stayed true to its Democrat partisanship and endorsed McBride.

The Times’ political bias appears to dictate the rulings handed out by its PolitiFact Florida project. A self-proclaimed political referee, PolitiFact Florida tilts its “Truth-o-Meter” heavily against Republican political figures. A Media Trackers Florida analysis shows PolitiFact Florida targets Republicans with 11 times more “Pants on Fire” rulings than Democrats.

According to the Media Trackers analysis, PolitiFact Florida issued its derisive “Pants on Fire” ruling against 27 individuals. Of those 27 individuals, 22 are Republicans, three are Independents or individuals with no clear party affiliation, and only two are Democrats.

 

[fskimospy]

I like Politifact and use it relatively often. Conservatives have complained about selection bias for years and I gave you evidence that shows that this may indeed be the case. When only 17% of liberals warrant a "pants on fire" rating vs 83% for conservatives (5 to 1 ratio), this infers one of three possibilities: (1) that liberals are 5 times less likely to say ridiculous lies as conservatives, (2) the site is disproportionately fact-checking conservatives and grading them more severely, or (3) both (1) and (2) are true to some degree. You chose. Believe what you want to believe, disregard the rest.

No. How many times do I have to explain this to you? The numbers you are talking about here are tiny and it is not possible to make those kind of judgments. Here's just a small sampling of the major flaws in what you're trying to do:

1. Your results are driven by outliers. Just two people comprise about a third of the total entries.

2. Many of the rulings are repeat rulings. People keep saying the same false thing and so one lie will rack up multiple entries. Highly problematic.

3. At the current time 'Pants on Fire' rulings comprise less than 10% of the site's total rulings. This means you are using an outlier laden dataset to evaluate unlikely events.

4. Total entries are about 100. It should be obvious why that's bad.

5. You continually fail to state WHAT YOU THINK THE APPROPRIATE DISTRIBUTION SHOULD BE. THIS IS AN ABSOLUTE REQUIREMENT.

I could go on, but I hope I've shown you how ridiculous you're being. Speaking of believing what you want to believe, I also find it funny that you are convinced of Politifact's partisanship but uncritically accept the analysis of "conservativefactcheck.com".

 

[realibrad]

I like Politifact and use it relatively often. Conservatives have complained about selection bias for years and I gave you evidence that shows that this may indeed be the case. When only 17% of liberals warrant a "pants on fire" rating vs 83% for conservatives (5 to 1 ratio), this infers one of three possibilities: (1) that liberals are 5 times less likely to say ridiculous lies as conservatives, (2) the site is disproportionately fact-checking conservatives and grading them more severely, or (3) both (1) and (2) are true to some degree. You chose. Believe what you want to believe, disregard the rest.

No, the error is in the 2nd part.

They may very well be disproportionately check republicans, but it does not mean they are grading them harder on what they do grade.

There are more than 2 options. It could also be true that conservatives are more than 5 times more likely, and that they fact check them less to get to the 5 times more. There are many options. You are empirically wrong.