Statistically speaking, this is a crock

I'm no authority on statistical theories and applications, but you can understand my skepticism when a supposedly serious research study, "A Measure of Media Bias," soon to appear in the Quarterly Journal of Economics, develops and utilizes a methodology that leads to some pretty freakish conclusions:

Of the 20 major media outlets studied, 18 scored left of center, with CBS' "Evening News," The New York Times and the Los Angeles Times ranking second, third and fourth most liberal behind the news pages of The Wall Street Journal.[emph added]

No, you read that correctly. The news pages of The Wall Street Journal stand as the most liberal news outlet in the country. FOX News' Special Report with Brit Hume is tagged as being reasonably centrist, being the fourth most centrist news outlet studied.

Also, the same methodology yields these jarring results: the National Rifle Association is a centrist group (you can pry my right-wing media bias out of my cold dead hands) while the Rand Corporation leans left, along with the Drudge Report.

Back to the study's press release:

The researchers took numerous steps to safeguard against bias — or the appearance of same — in the work, which took close to three years to complete. They went to great lengths to ensure that as many research assistants supported Democratic candidate Al Gore in the 2000 election as supported President George Bush. They also sought no outside funding, a rarity in scholarly research.

Perhaps it's best to stick with the "appearance of same" claim.

"No matter the results, we feared our findings would've been suspect if we'd received support from any group that could be perceived as right- or left-leaning, so we consciously decided to fund this project only with our own salaries and research funds that our own universities provided," [study author] Groseclose said.

Are you impressed with the purity of all involved? Don't be.

The study's lead author is Tim Groseclose, a UCLA political scientist; its co-author is Jeffrey Milyo, University of Missouri economist and public policy scholar.

And even if Groseclose and Milyo funded this study with their own loose change, they are not political virgins. Sid's Fishbowl documents the whoring.

According to the professor's curriculum vita, [Groseclose has] received the following "honors and fellowships":

 Hoover National Fellow
 Olin Faculty Fellow
 Lambe Fellow, Institute for Humane Studies
 Dissertation Fellow, Center for the Study of Public Choice, George Mason University

Co-author Jeffrey Milyo was a Salvatori fellow for the ultra-right-wing (by their paper's own numbers!) Heritage Foundation. He and Groseclose wrote their first article together in 1996 for the far-right scandal sheet The American Spectator.

This is astonishing. So two wingnuts whose work has appeared in the excreble Specator are now being published in The Quarterly Journal of Economics. The QJE is no rag. It's the oldest professional journal of economics in the English language, and is edited at Harvard University's Department of Economics.

What the hell.

UPDATE Media Matters looks at how the media are reporting on the media bias study — and, of course, failing to mention the authors' involvement with right-wing groups.