Almost two years ago, former governor of Alaska Sarah Palin sent out her infamous “death panels” post on Facebook. The media couldn’t resist the incendiary update. Soon, the two-word phrase—a demonstrable misrepresentation of the end-of-life counseling options proposed in the president’s health care plan—turned up the heat at Town Hall meetings and lodged itself in the discourse like a tough string of beef between two closely-packed molars.

How did this colorful lie wedge its way so far in? Dr. Regina Lawrence of Louisiana State University’s Manship School of Mass Communications asks that very question in a new study co-authored by Matt Schafer, which you can read about in Schafer’s article, “Two Years Later: The Media Response to Death Panels and Why It’s Still Important.” For the report—to be presented at this weekend’s International Communications Association Conference in Boston—Lawrence and Schafer analyzed over 700 newspaper stories from the country’s top fifty newspapers, and a number of network news reports. The pair sorted the stories that debunked the claim from those that did not, and then noted those that dutifully included quotes from all sides, probing just how effective the media was at getting to—and sometimes getting around—the facts. Here, Lawrence discusses her report with CJR assistant editor Joel Meares.

Seven hundred stories. On a Sarah Palin quote. How long did it take you and Matt to complete the study?

It took longer than we anticipated because the first time we went about coding all of these news stories we went in with the assumption that the stories would either debunk the claim or they would present the claim in a he-said/she-said manner. Because we went in with that assumption, our coding scheme didn’t have a way of handling stories that did both. That led to difficult judgment calls we weren’t comfortable with. After doing all of that we finally realized that, “Wow, this is actually a really interesting finding,” but we need to go back and redo the entire sample, which took months.

What hypothesis were you trying to test with the original coding scheme?

It was not so much hypothesis-testing as exploring a research question. We didn’t go in—to put it in scientific terms—with “directional expectations.” We just wanted to see how journalists treated this claim. Knowing that it had been debunked early and fast by PolitiFact and, we were very curious how journalists made use of that information and how they treated the claims.

Why did you decide to look at the death panels story as opposed to other stories involving misinformation that had proliferated?

Part of it is simply timeliness. This whole project came out of a class I was teaching that my co-author Matt Schafer was in. I had made some sort of offhand comment, like, “Oh I bet even death panels are getting covered in some certain kind of way…” It was just talking off the top of my head, but Matt went out and gathered a small sample of articles and then came to my office and said, “Look at this, I think I have something interesting here.” This would have been the fall right after the summer of the raucous Town Hall meetings, so it was timeliness that drew us to it.

But on a deeper level, it’s a little bit easier than doing something like climate change, which is a very big and complicated issue—you can’t tie it down to a single claim very easily. Here, you have a case of misinformation that was crystallized into a single catchphrase.

Do you see parallels between the way that the death panels story got out there and misinformation about climate change?

Absolutely. You’ve got pretty organized message machines on both sides of the issue, but I think a very organized message machine on the more right-leaning side of the media and political aisle. And you’ve got a public that is increasingly polarized. That’s not just an offhand observation, that’s something that a lot of research is showing: people are much more innately dividing into camps that could be defined as more liberal or more conservative on these kinds of issues. It’s almost like these issues become shortcuts for people to have their predispositions activated. Say “climate change” and people start to sort themselves into one side of the room or the other. Based on the available data, I think the same is true with death panels. If you say that word in a lot of rooms full of people, I think you would see a fair amount of sorting going on.

Why did you decide not to look at cable news? That would seem to be where a lot of death panel coverage was happening.

Joel Meares is a former CJR assistant editor.