UCLA, Stanford Study shows Drudge Report most centrist of media outlets

[G4scott]

link (http://216.239.51.104/search?q=cache:tCEKSNVW-OYJ:mason.gmu.edu/~atabarro/MediaBias.doc+drudge%3F&hl=en)

Who would've thought?

For those who think the US media is part of the Vast Right-Wing Conspiracy, think again...

[zimv20]

heh heh, i read some of it. the method they use to determine that is interesting, to say the least. also of interest is their finding that the ACLU is more conservative than the average GOP member of congress.

g4scott - thanks for deciding not only what reports i need to to pay attention to, but what conclusions i need to draw. i don't know what i would do if i acutally had a brain of my own.

[pooky]

Ok, first off, your title is way off. UCLA and Stanford didn't find this, a couple of graduate students wrote a paper. Usually when people cite a "Stanford Study," for example, it's in a peer reviewed journal. That this article is presented as a .doc on someone's personal website makes it highly suspect. It also reads like a high school term paper, which makes me think it might be a joke, or at the very least something that didn't have a lot of time devoted to it. Finally, their methods are very suspect; I've never heard anything like this. I'd like to hear a real political scientist voice an opinion on this one.

[Neserk]

*busted*

[wordmunger]

Ok, first off, your title is way off. UCLA and Stanford didn't find this, a couple of graduate students wrote a paper. Usually when people cite a "Stanford Study," for example, it's in a peer reviewed journal. That this article is presented as a .doc on someone's personal website makes it highly suspect. It also reads like a high school term paper, which makes me think it might be a joke, or at the very least something that didn't have a lot of time devoted to it. Finally, their methods are very suspect; I've never heard anything like this. I'd like to hear a real political scientist voice an opinion on this one.
Actually, Jeff Milyo is a professor of public policy at the University of Chicago (link (http://chronicle.uchicago.edu/001102/milyo-q-a.shtml)) and Tim Groseclose is a professor of political economy at Stanford (link (http://faculty-gsb.stanford.edu/groseclose/homepage.htm)).

But even if they're professors, if this is an unpublished manuscript it means it hasn't stood up to peer review. Noam Chomsky is also a "professor," but I think you'd find his opinions diametrically opposed to Groseclose and Milyo.

[wordmunger]

Okay, now I've skimmed the article, and I have to say it's an appealing approach to the problem. Basically what they do is analyze which think tanks the media outlets cite, and compare them to which think tanks are cited by conservative and liberal congressmen. So if liberal congressmen cite a think tank more often than conservative congressmen, then that think tank is deemed "liberal," and media outlets which disproportionately cite that think tank are also deemed liberal.

But what if liberal think tanks are cited more often by the media because they do better research? Or because they produce more research? It seems to me that an economic analysis of media sources misses the point. Also, what if conservative media simply don't cite think tanks as often, period? Then it will look like the "liberal" media are disproportionately citing "liberal" think tanks, because they cite them more often than the "conservative" media.

[IJ Reilly]

Does the study consider the context of the think-tank citations? IOW, if I write in a story, "according to those fools at the Cato Institute...", does that make me Libertarian-leaning?

[zimv20]

i may have heard it here, or maybe not...

a study was done on NPR's citing of think tanks. of the think tanks they cited, an overwhelming number of them are considered conservative.

edit: here (http://www.fair.org/extra/0405/npr-study.html) it is...


FAIR’s four-month study of NPR in 1993 found 10 think tanks that were cited twice or more. In a new four-month study (5/03–8/03), the list of think tanks cited two or more times has grown to 17, accounting for 133 appearances.

FAIR classified each think tank by ideological orientation as either centrist, right of center or left of center. Representatives of think tanks to the right of center outnumbered those to the left of center by more than four to one: 62 appearances to 15. Centrist think tanks provided sources for 56 appearances.

The most often quoted think tank was the centrist Brookings Institution, quoted 31 times; it was also the most quoted think tank in 1993. It was followed by 19 appearances by the conservative Center for Strategic and International Studies and 17 by the centrist Council on Foreign Relations. The most frequently cited left-of-center organization was the Urban Institute, with eight appearances.

[wordmunger]

Does the study consider the context of the think-tank citations? IOW, if I write in a story, "according to those fools at the Cato Institute...", does that make me Libertarian-leaning?
Yeah, they cover that. They don't count citations like that. They also hired 50 percent Bush supporters and 50 percent Gore supporters as research assistants.

[Taft]

I had a discussion about this on SpinSanity.org a few weeks back. Its in the first article's comments on this page (http://www.spinsanity.org/post.html?2004_05_16_archive.html).

Basically, I found the following issues with the study:

What percentage of common news stories cite a think tank? I believe it to be a tiny fraction of total news stories. Further, I don't believe those stories are at all representative of the average news story.

The study also does not even attempt to justify the use of these types of stories as a representative of all news stories. This should be an important point they should show evidence for. Also, your assertion that the fact that the majority of stories citing a think tank leaning one way or the other as an indication of the whole publication's bias is specious. The author of the study also uses this assumption, but again fails to adequately justify this position in his study.

Given that:

1) The news samples used are not (in my opinion) representative of the average news story.
2) The author gives little justification for his choice of samples and why those samples would be representative of the total population.

I wouldn't take this study seriously unless it were peer reviewed or I could get a better understanding of why the author thinks his underlying assumptions are true. And that is my major problem with the study: the author assumes without providing justification for those assumptions.

I see where he was going, I just totally disagree with his metric being indicative of institutional media bias.

Taft

[Taft]

Does the study consider the context of the think-tank citations? IOW, if I write in a story, "according to those fools at the Cato Institute...", does that make me Libertarian-leaning?

No it does not. It also does not attempt to justify the use of ONLY news stories which make citations as a representative of ALL news stories.

I think measures like this are pretty seriously flawed. There are far too many unaccounted for factors to make this study useful as a general indicator of bias.

Taft

[wordmunger]

No it does not. It also does not attempt to justify the use of ONLY news stories which make citations as a representative of ALL news stories.

I think measures like this are pretty seriously flawed. There are far too many unaccounted for factors to make this study useful as a general indicator of bias.

Taft

I agree with your conclusion, but you are wrong when you say the study does not account for situations when the think tank was quoted only to demonstrate that it is wrong. Here's a direct quote from the article:

*Also, we omitted the instances where the member of Congress or journalist only cited the think tank so he or she could criticize it or explain why it was wrong.* About five percent of the congressional citations and about one percent of the media citations fell into this category.

[pooky]

I'd also have to add (and this is admittedly personal bias) that ANY study that reveals Fox News as a "centrist" source is bunk. Period. Just watching Fox News for any length of time will tell you it is far from centrist.

[Neserk]

I'd also have to add (and this is admittedly personal bias) that ANY study that reveals Fox News as a "centrist" source is bunk. Period. Just watching Fox News for any length of time will tell you it is far from centrist.


I saw a few mintues of it on Sunday morning :lo:

"And now for news you won't see on any other station, the good we are doing in Iraq"

That is almost a direct quote! I had to laugh. Then I changed the channel.

[Taft]

I agree with your conclusion, but you are wrong when you say the study does not account for situations when the think tank was quoted only to demonstrate that it is wrong. Here's a direct quote from the article:

That is true.

I guess I should clarify, however. What I meant is that a liberal senator could use a statistic which, out of the context of the original think tank writing, could be used for purposes other than the original writing intended.

Lets make up a scenario (please remember that all of the following IS MADE UP). Lets say the Cato institute put out a study which, in a small part of the study said, "15% of homes which own guns eventually will have an accident." They might have intended that statistic to aide a larger argument or point. However, a liberal senator might be doing research in order to argue against a different conservative senator and happen upon that statistic. Then, using that statistic out of context, and applied to a different argument or conversation, the liberal senator might cite the study.

What I'm saying is that in that scenario, the fact that the liberal senator cited the study does not in any way endorse or refute the overall findings of the study. Rather, the study was dissected for information helpful to that senator in a different context.

So far as I can tell, this type of citation would be used in their sample. I am not, however, in any way sure of how many such citations might occur. It may be a rare case.

Taft

[wordmunger]

So far as I can tell, this type of citation would be used in their sample. I am not, however, in any way sure of how many such citations might occur. It may be a rare case.

Taft

Yes, excellent point. I was trying to make the same point above, but I wasn't able to do it with your kind of clarity. The most important point is that this is an economic study, not a rhetorical analysis. As an economic study, it attempts to avoid all possible bias, but it can't avoid the most substantial bias of all, which is that there is no way for an economic study to capture all the nuances in a survey of thousands of documents. Unfortunately, neither can a rhetorical analysis--because it will introduce its own sorts of biases.

[Taft]

The most important point is that this is an economic study, not a rhetorical analysis. As an economic study, it attempts to avoid all possible bias, but it can't avoid the most substantial bias of all, which is that there is no way for an economic study to capture all the nuances in a survey of thousands of documents. Unfortunately, neither can a rhetorical analysis--because it will introduce its own sorts of biases.

Good points. I do think, however, that an economic study could be created which would do a [i]better[/b] job at quantifying bias through specific measures, even if the study wasn't "perfect" (what study is?). I just think the SINGLE measure this study relies on is far too simplistic to be of any use. I would be interested in seeing a study which approaches the problem in the same way, but uses more complete and meaningful measures.

Taft