By now, you’ve no doubt heard about the study claiming Facebook will go belly up in a few short years. The evidence, it would seem, lies in Google search trends. Those data show a precipitous drop in interest in the world’s most popular social network. And that decline, the reasoning goes, foretells an impending mass exodus from Facebook.

If this all sounds a bit dubious, there’s good reason. The prediction of Facebook’s imminent demise is based on a lot of assumptions, among them the troubling (but common) error of passing correlation off as causation.

Even Facebook weighed in. Their whimsical post raised concerns about the future supply of air based – you guessed it – on Google search trends as a predictor.

There are other problems with the study, which you can review for yourself. Besides conflating causation with correlation, it fails to define key terms like “decline,” and a host of typos raise doubt about the rigor of the statistical analysis employed. Put simply, it’s all a bit sloppy.

Kind of like some of the criticisms of the paper.


You may wonder: Could such a questionable study really emanate from one of the word’s most prestigious research universities?

Is this really the product of Princeton researchers?

If you’ve followed the coverage, the answer is clearly “yes.” The Guardian calls the authors “Princeton researchers.” Many others do, too. Mashable calls them “scientists.” On CNN, they were a “group of Princeton researchers.”

I don’t know about you, but to me it sounds like they’re salaried professionals, probably faculty.

But, as Reed Albergotti made clear on the Wall Street Journal’s Digits blog, the authors are, in fact, a pair of Ph.D. students.

Is it really accurate to say they’re Princeton researchers? I think it’s a stretch. And certainly misleading.

But some of the criticism I’ve seen is looser still, attributing the paper merely to Princeton itself, as if these findings somehow reflect the institution’s official stance.

Reading all the incredulity over the bold prediction, you might conclude that the study has been published in a peer-reviewed research journal, sanctioned by academe at large.


Albergotti made that clear, too, pointing out that the authors “were withholding comment until their paper is peer reviewed.”

Will Oremus at Slate also underscored that the authors are grad students, their work published only in that it’s been made a PDF and posted to the Web.

In other words, a couple of Ph.D. students who happen to study at Princeton uploaded a PDF of something they’ve been working on.


I get it. This is a much more interesting story if we pin this questionable research on the prestigious Princeton. If we make it sound like a big, funded research project, all the better.

But, in doing so, aren’t we oversimplifying things—not unlike the researchers themselves? Wouldn’t it be more intellectually honest if we framed this “research” within the right context?

The trouble with that approach, of course, is that it takes the air out of any controversy. Because “a draft student paper” just isn’t as enticing as “research from Princeton.”

But it is more accurate.

The trouble with this story isn’t just passing the paper off as bonafide research. It’s making it a story in the first place.

My guess (an assumption, admittedly) is that the students who wrote this paper are, like most Ph.D. students, learning how to conduct research. As they should be. Will mistakes be made in that process? Absolutely. Are they taken aback by the uproar over their work? I think so.

No doubt both will learn a lot. They’ll learn how to get better at the scientific method. And they’ll learn how Twitter, blogs and news outlets sometimes fall into the trap of making a tempest in a teapot.

But the real lessons here apply to all of us, whether we’re about to hit publish on a post, click to retweet or simply find ourselves awash in the day’s attention-grabbing headlines.

Sometimes, in the rush to grab eyeballs and in the desire to learn about something interesting, accuracy suffers.

Email this to someoneShare on FacebookShare on Google+Share on LinkedInShare on RedditShare on TumblrTweet about this on Twitter