You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.

Do 11 Percent of Americans Really Think HTML Is an STD?

The strange saga of a possibly real, possibly fake study that became an Internet news meme

Shutterstock

On Monday, veteran Washington Post editor and New Yorker contributor Marc Fisher published a deeply reported, scrupulous Columbia Journalism Review cover story on how the Internet’s metabolism and economy, which places a premium on being first to a story and on attracting clicks, has led to compromises when it comes to the whole accuracy thing. With surprisingly little sorrow or anger, Fisher concluded, “The imperative for speed in the journalism of tweets and Vines has triumphed over traditional ways: What’s news is what’s out there, whether or not it’s been checked and verified. Rare is the news organization that doesn’t occasionally jump on Twitter with half-baked facts.”

As if on cue, a fun news story has been making the rounds in the past few days: A survey found that 11 percent of Americans believe that "HTML" is a sexually transmitted disease. Other findings included that 20 percent believe a "motherboard" is a cruise-ship deck and 15 percent believe "software" is a type of clothing. Old media like Time and The Los Angeles Times picked it up—in fact, the LAT appears to have been Patient Zero. Middle-aged media (I mean that in a good way) like Talking Points Memo and Jezebel posted it. And, of course, it wouldn’t be a trend story about changing news mores without a BuzzFeed mention: The site posted the story as well, making its sourcing look better than it was by introducing “a new study from Vouchercloud.net,” with “Vouchercloud.net” hyperlinked to that company’s homepage and “study” hyperlinked to … the LAT article. 

The survey itself may not exist, though. The website iMediaEthics reported Tuesday night that the LAT, which hadn’t linked to the survey, also hadn’t seen the survey—just a press release “reporting” the survey that was disseminated by a publicity firm that represents a British coupon company, Vouchercloud. iMediaEthics then learned that the P.R. firm, 10 Yetis, lists viral marketing and stunts as among the things it does. As of Wednesday around noon, neither iMediaEthics nor any of the people who reported on the survey had actually seen the survey itself.

My point isn’t to smirk—except, perhaps, at those math-challenged reporters and editors who believe 11 percent is closer to 1 in 10 than it is to 1 in 9. (Jezebel's headline is truer than it realizes: "1 in 10 Americans Believe HTML is an STD. LOL. We're Dummies.") Rather, as Fisher notes, this is undeniably the world we live in: one where monetizable clicks incentivize posting catchy content, and posting it quickly, over principles like accuracy and newsworthiness.

Williams Pelegrin, the author who posted at Digital Trends, replied to my request for comment. “I originally went to Reddit, through which I clicked through to the article that was on The LA Times,” he emailed. “I did find the story funny in that it seemed more like a joke rather than something that was legit, and I tried to see if anyone else reported on anything similar. I didn't come across any, and since I don't really assume they would just get some phony information and try to turn it into something that would grab headlines, I took it at face value and reported on it.” He thanked me for bringing this to his attention, and informed me he would update the post immediately. (Pelegrin will issue the correction, but there’s an old saying that pre-dates the Internet: “A lie can travel halfway around the world faster than the truth can put its pants on.”)

What are the lessons here? In Pelegrin’s case, he trusted the Los Angeles Times—and why not? It’s a storied, legacy newspaper—even if Sam Zell did his best to destroy it—but even still, he got burned. It’s not so much that the center cannot hold as that there is no center anymore. So don't trust anybody, not even ancient papers with Times in their titles.

The main culprit here—speed—is perhaps the solution, too. It is telling that BuzzFeed, legendary for its army of writers and total dedication to being the most popular kid on the Internet, was among the first (perhaps the very first) to post an update acknowledging the story’s dubiousness. The same metabolism that led BuzzFeed to post quickly in the first place also led it to update quickly. Then again, the update itself is rather milquetoast—"Update: Some have called into question the validity of this study."—and appears at the very bottom of the article, which otherwise remains unchanged and continues to rack up social shares. (As of this writing, TPM and Time had posted updates, Jezebel and LAT had not.)

Fisher suggests that we will find an equilibrium, as outlets—and advertisers—see the benefit not only of clicks (and therefore speed) but of brand credibility (and therefore of accuracy and quality content). We are seeing this shift play out in the design realm, in which publishers are moving away from click-happy pagination and toward reader-friendly, single-page scrolling. I do not think we are yet seeing it in the realm of posting shareable content.

I do think some salvation may lie in professional pride. It’s salutary that culture has not caught up with technological change, at least not yet. According to his Google Plus page, Pelegrin graduated from college just last year. He has not known a journalistic landscape without the Internet; he may not have even known one without Twitter! Yet he wrote me, “I’d much rather be honest and transparent with this stuff than leaving it as it is,” which sounds like what a journalist 30 years his senior might have said. Journalistic culture still cares about these things, in other words, even if it might be better for their business models if they didn’t. That may save us a lot of future hoaxes.

And save a lot of unnecessary confusion, too: Just as I was about to publish this piece, I noticed that Time had posted two more updates on its article.  10 Yetis, Vouchercloud's P.R. firm, had responded to a request for comment, telling the magazine, "It’s a completely real study and survey that we sent out to a large database. I’ve read the claims I know exactly what you’re talking about, but it is a real tried and tested method. We’re quite well known in the U.K. for doing these. It’s 100% genuine and what we do is 100% valid." The story doesn't end there, of course; 10 Yetis's response only complicates it further. I have reached out to both Vouchercloud and 10 Yetis, and yes, I will most certainly update this post if I hear back from them.

UPDATE: 10 Yetis emailed me the survey results, along with the following explanation: "It was just a fun survey sent by email to an opt-in database of American nationals, all over 18 yrs old. I can confirm that it's not a publicity stunt & the results are 100% real." It sounds like the type of survey in whose results I would not place too much stock, in other words, though maybe I'm being cynical. Then again, maybe cynicism is a hallmark of sound journalism.

Image via Shutterstock.