Friday, March 14, 2008

Foxtacular Friday fibfest

Here's a real gem from the Fox front page. Indeed, given all the fun cultural variables in play, it's kind of surprising that it's not the lede story:

Study: Antiwar Reporting Helps U.S. Enemies
Insurgents in Iraq get a boost from coverage in the news media that shows support for troop withdrawals from the war torn country, according to a study.

Two Harvard University economists found that insurgent groups are responsive to "antiresolve" statements in the media.


See all the Science Magic? It's a study! It's from Harvard! It's by economists (so it's from the clear-eyed realistic perspective)! We're not going to get a lot out of the five grafs that make up a standard Fox rewrite, though, so we have to go to the source -- well, sort of. Fox cribbed this one from US News and World Report:

Are Iraqi Insurgents Emboldened by Antiwar Reporting?
Economists say their study, with caveats, finds some linkages
Are insurgents in Iraq emboldened by voices in the news media expressing dissent or calling for troop withdrawals from Iraq? The short answer, according to a pair of Harvard economists, is yes.

If our friends over at Language Log are wondering how conventional wisdom about "science" gets so entrenched, this is how. Fox isn't just pointing to a "study." It's pointing to what somebody else has already written about what the "Harvard economists" found. No need to link to the paper when you're quoting the people who link to a place where you can find the abstract!

In a paper published by the National Bureau of Economic Research, the authors are quick to point out numerous caveats to their findings, based on data from mid-2003 through late 2007.

Really? Did you find some caveats before page 20?

Yet, their results show that insurgent groups are not devoid of reason and unresponsive to outside pressures and stimuli.

Watch the pivot foot. US News is signaling that there's a second discovery besides the one asserted in the lede, and it's designed to bolster the study's credibility by making it look "objective": See? These counterintuitive dudes have found that insurgents "are not devoid of reason"! Which the rest of the world has known since -- well, roughly since Josephus Flavius turned in his first draft of "The Jewish Wars." News apparently travels slowly to Cambridge.

If you haven't yet had your adult minimum daily requirement of amateur international expertise, go read the USNWR piece in all its glory. The point is basically the same: Every time those feckless news media cast doubt on The Mission, they're killing Americans and Iraqis! But if you're a content analysis fan, you're probably wondering: Interesting conceptual definition of the independent variable there ("coverage ... that shows support for troop withdrawals" or "voices ... expressing dissent"). Wonder how that's going to be operationalized?

Here's how. We're going to measure "emboldening" statements, which are going to send signals about US resolve, by counting the number of times "top Bush administration officials ... refer to statements or actions by other U.S. political figures that might encourage violent extremist groups in Iraq." In other words, the turrists aren't emboldened when the Democrats say their usual cut-n-run stuff; they're emboldened when the White House press secretary says the Democrats are the party of cutting and running.

How's the intercoder reliability on that variable? There isn't any! The Harvard economists used an iterative computer program that searches a database,* and everybody knows that computers are never wrong.

There's no substitute -- ever -- for asking for the original data, whether the subject is "Poll says X!" or "MRI study shows gals talk more than guys!" or "Craven liberal weenies are doing our enemies' work for them!" Ask the reporter (or call the news agency): What did these guys measure? How did they measure it? What did the results say? What stuff in the real world stands in for the concepts they're aiming for? By the time it becomes "study says," it's on its way to being conventional wisdom. The editor's job is to hold up that train until all the boxes are checked.

And when your readers call to demand why you haven't run that Fox story that proves how all you media liberals cover up the truth? That's when you get to say, "Well, Fox lied about the results of a study that measured something else entirely. Would you like to stop by and read the thing?"


* You can get to an explanation of the search for free here: http://www.nber.org/data-appendix/w13839/Iraq_Appendices.doc

Labels:

0 Comments:

Post a Comment

Links to this post:

Create a Link

<< Home