[Update: Craig Silverman elaborates on this column in a new CJR podcast, which you can listen to elsewhere on CJR.org here, or via iTunes here.]

Time for a pop quiz: How many of the leaked diplomatic cables in WikiLeaks’s possession has the organization released publicly?

A) Roughly 2,000

B) Roughly 250,000

C) None. They’ve all been released by media outlets.

I’m willing to bet that many people will get this wrong. Maybe even most people. Journalists certainly have been getting it wrong, which means the public has been fed a diet of inaccurate information for some time.

The correct answer is A: Roughly 2,000. But many news outlets continue to report that WikiLeaks dumped all 250,000 or so diplomatic cables online. This incorrect fact has spread far and wide. It’s also frequently cited as a reason why WikiLeaks does not deserve recognition or protection as a journalism organization.

In just one example, the president of the Society of Professional Journalists cited the incorrect figure in a blog post. Here’s how Hagit Limor’s post originally began:

If you’re looking for consensus on WikiLeaks, don’t ask a group of journalists. Several of our committees have been batting around the ramifications all week, and we can’t even agree on the most basic question: Is WikiLeaks journalism?

Those who say “no” call WikiLeaks a source, a conduit, a whistleblower. They call the 250,000 diplomatic cables posted online a data-dump without filters, fact-checking or context from other sources. They say there’s no original reporting, hence the need for established media partners to get out the word.

After being alerted to the mistaken figure, the correct “2,000” was inserted where “250,000” used to be—though no correction was added to the post in order to disclose the error and note the change.

I asked Limor how she reacted when she realized her initial figure was incorrect.

“My reaction was immediate: to set the record straight,” she said by e-mail. “Accuracy is topmost for any journalist. Our Society of Professional Journalists Code of Ethics places it as our first test under our tenet to ‘Seek Truth and Report it.’”

Further down in the SPJ’s Code, it also says that journalists must “Admit mistakes and correct them promptly.” As of this writing, the post is still without a correction/admission, a fact that was noted by a comment on the post, in addition to being pointed out to Limor by me in two e-mails. I’ll ask again publicly: please adhere to the SPJ Code of Ethics and add a correction to the post, noting the original error and when the fix was made.

NYU journalism professor Jay Rosen was among those who pointed out the error, though his comment was held for close to twenty-four hours before being approved. (It went live after I asked Limor to look into the issue. She explained that Rosen’s comment was inadvertently caught up in a spam queue.) When I pointed out the post’s lack of a correction, Rosen responded with this tweet:

One obvious concern about the prevalence of this mistaken number is that it could cloud the way people view WikiLeaks. If you’re under the impression that WikiLeaks indiscriminately dumped 250,000 diplomatic cables online without any kind of control or vetting on its part or that of its media partners, then you’ll probably have a different view than if you’re aware that it has publicly released roughly 2,000 cables, many of which have been vetted (and had parts redacted) by established media partners. (Salon’s Glenn Greenwald offers more detail about the vetting process used for the first 960 cables put online by WikiLeaks.)

Rosen views the number confusion as a significant error that has tainted the discussion about WikiLeaks.

“[Limor’s] corrected post now suggests that SPJ members skeptical of Wikileaks were skeptical of the organization dumping 2,000 documents willy nilly onto the web,” Rosen told me. “But clearly, what they had debated was the false premise: indiscriminate dumping of 250,000 documents. Is she claiming she went back and asked for a do-over of that debate based on the now corrected premise? No. She’s saying something that distorts the prior history.”

Limor doesn’t feel the incorrect figure changed the nature of the discussion within the SPJ.

“I think the debate has centered on whether this constitutes journalism, and that would have ensued no matter the number of cables so long as some were incendiary enough to garner interest,” she wrote. “While some journalists did refer to the incorrect number as their reason to exclude this as journalism, citing it as a data-dump, I believe they would have said the same for almost 2,000 cables. However, as noted in my first blog on the subject, some journalists clearly found the portion of the WikiLeaks website that claims to verify material and remove details that would harm ‘innocent people’ and they cited that in defense of WL, even for the higher, incorrect number.”

This situation highlights how a seemingly simple matter of an incorrect number can significantly alter one’s understanding of an issue. Two thousand mostly-vetted cables are very different than 250,000 unvetted, dumped cables. Yet the material change is the matter of just a few digits.

This error was initially brought to my attention thanks to a blog post by NPR ombudsman Alicia Shepard. A persistent NPR listener repeatedly contacted the organization to object to its reports that said WikiLeaks had publicly released all 250,000 cables.

Here’s the correction published by NPR on December 28:

In recent weeks, NPR hosts, reporters and guests have incorrectly said or implied that WikiLeaks recently has disclosed or released roughly 250,000 U.S. diplomatic cables. Although the website has vowed to publish “251,287 leaked United States embassy cables,” as of Dec. 28, 2010, only 1,942 of the cables had been released.

Along with Greenwald and Shepard, Matthew Schafer, a graduate student at Louisiana State University who writes the Lippmann Would Roll blog, has also been on the case.* He recently provided a list of news organizations that have either been vague or completely incorrect when reporting the figure. His post lists questionable examples from news organizations including the New York Daily News, UPI, the Christian Science Monitor, and the AP. (In fact, the SPJ’s Limor told me she was led to the wrong number by an AP report.)

“The following examples are just a few instances of lazy journalism,” Schafer wrote. “Some illustrate ambiguous language, while others’ language is just flat out wrong.”

I’d add another rather surprising culprit to his list. Der Spiegel, the German weekly with a huge fact checking department, is among the media outlets to receive leaks directly from WikiLeaks. Yet a summary on this story reports that “The US, for its part, would like to try [Assange] for making 250,000 confidential diplomatic cables public.”

To be fair, other news organizations have been careful about how they describe the WikiLeaks documents. Greg Brock, a senior editor who oversees corrections at The New York Times, told me the paper is in possession of all 250,000 cables, which means it’s technically correct if it says the documents have been released (as in released to the Times). He also pointed to several examples where the paper was delicate in its descriptions of the documents, such as this:

But with the initial series of articles and cable postings nearing an end, the fate of the roughly 250,000 cables that have not been placed online is uncertain. The five publications have announced no plans to make public all the documents. WikiLeaks’s intentions remain unclear.

Or this:

[Assange’s] incarceration has not stanched the controversial flow of classified American documents from WikiLeaks, the most recent drawn from some 250,000 diplomatic cables, mostly between American diplomats abroad and the State Department in Washington.

That said, Brock agrees the language being used by many media outlets is problematic, if not wholly incorrect.

“I think you’re correct that the language being used causes the confusion,” he said in an e-mail. “‘Released’ means different things to different readers. But I think the average person would take that to mean released ‘publicly.’ But they did ‘release’ them to several news organizations.”

Hopefully, the NPR correction and the attention it has garnered will lead to other, similar offerings from news organizations—and put an end to the mistaken reporting. (Politico, for example, corrected a report thanks to Schafer’s work on the matter.) In that respect, this correction from the January 6th edition of The Guardian is good news:

Accompanying a story about confidential US diplomatic cables leaked to the whistleblowing website WikiLeaks, a panel - HaikuLeaks, 31 December, page 19 - began: “For those with neither the time nor inclination to wade through all 251,287 of the diplomatic cables released by WikiLeaks…”. So far, WikiLeaks has published only a portion of this total.

Correction of the Week

A PHOTOGRAPH that accompanied a story in yesterday’s Daily News was not of the massage therapist Shannon O’Toole, but of a different woman, an author also named Shannon O’Toole, who wrote ‘Wedded to the Game: The Real Lives of NFL Women.’

“That book, published by the University of Nebraska press in 2006, is a nonfiction account described in reviews as ‘realistic’ and based on interviews and surveys with wives and girlfriends of NFL players and coaches. She is not involved in a lawsuit massage therapists Shannon O’Toole and Christina Scavo filed against the Jets on Monday.” - New York Daily News

*Correction: This article originally misspelled the name of a Louisiana State grad student who maintains a blog called Lippmann Would Roll. He is Matthew Schafer, not Shafer. The misspelling has been corrected. CJR regrets the error.