the water cooler

David House on Post-Publication Fact-Checking and Deterring Plagiarism

October 7, 2005
David House

In March of 2004, the editors at the Fort Worth Star-Telegram announced that they would begin implementing a new deterrence system designed to safeguard their paper in the future from the likes of fabulists like Jayson Blair and Jack Kelley. David House, the reader advocate at the Star-Telegram, was put in charge of the system, which takes randomly selected articles that have already been published and subjects the stories to intense factual scrutiny. “As far as can be determined,” House wrote at the time, “no other newspaper in the United States has a similar policy.” So how has the experiment worked over the past year-and-a-half? We recently caught up with House to find out.

Felix Gillette: How do you choose which articles to comb through in search of malfeasance?

David House: Basically, we want the candidate story for the exercise to be selected at random. I begin with a stack of a week of papers. All three local sections are included. I take those papers and shuffle them like a deck of cards. Then I take some of the sections in some of the papers and shuffle them around. Then I take this big, black, Sharpie, and start going through each paper from page one on. I mark local stories that are good candidates for the exercise –which is to say, stories that are longer than eight inches and have some degree of enterprise or difficulty, or some characteristic that presented a challenge to the reporter.

I number them all the way through a week of papers. Then I take the lid to my box of stationary in which I keep a hundred-and-some-odd red movie tickets. On the back of each ticket I’ve written a number from 1 to 135. I have those tickets in the box lid. Then I call the executive editor Jim Witt and ask him if he would like to do a drawing. So I go in there, and I hold the box lid above his head. He reaches up. He can’t see the tickets. He stirs them around and pulls one out. The number that’s on the back of that ticket corresponds with the number of the story that’s going to be fact-checked.

FG: Just one?

DH: Just one, because most of the time it takes us a while. The news researcher who works with me really has a full plate. She works in the fact-checking as she can. For a while, we were using iThenticate … a plagiarism detection program [from a company] based in Oakland, California. They have this incredible database and Web crawler. Basically what we were able to do with iThenticate was run a story through their database and it would go through millions of documents and find matching phrases. It would subject the story to all sorts of scrutiny and send you a report back. Meanwhile, the news researcher would be checking the story against the Lexis-Nexis database and the Web. And I get help from D’Juana Gibson, who is the executive editor’s administrative assistant, who helps me call the sources in the story to ask them whether they were quoted or paraphrased accurately and whether they have any issues with the story.

Sign up for CJR's daily email

FG: Have you caught any acts of plagiarism?

DH: No.

FG: Have any sources complained that they were treated unfairly?

DH: Basically, we have gotten a consistently high grade from sources. There was one instance in which we fact-checked a lengthy piece in the features section about a performing arts center. There was a lot of conflict internally at this arts group. The story was fact-checked, and I also sent accuracy letters to the sources. It gave them an opportunity to vent.

But almost without fail, the folks are just really impressed that we called. They don’t expect us to do that. They don’t ever expect to hear back from anyone in the news media. So when they get a call from us, or an accuracy letter, they say thanks for caring, we didn’t think you did. All of a sudden a big, faceless institution is warm-blooded to them, and human.

FG: How many corrections do you tend to turn up as a result of these post-publication fact-checks?

DH: They don’t necessarily pile up any inaccurate information. But we don’t hesitate to run corrections or clarifications.

FG: So the system hasn’t significantly increased the number of corrections you have to run?

DH: No. It’s been really interesting in that regard. We figured, if nothing else, maybe we could flush out some errors or something. But it just hasn’t happened.

FG: So how do you know if the system is working?

DH: By zero incidences of plagiarism. Maybe that would be the case anyway. But that’s what we’re most interested in. We also know — and we don’t have any formal measure of this — we just know that among news staff there’s always a brief buzz when I send an email to a staffer saying your story has been selected for the fact-checking exercise.

FG: How have the reporters reacted?

DH: They were apprehensive about it at first, as we knew they would be. But now they’re fine about it.

FG: Do you give them any feedback afterwards?

DH: Oh, yeah. I send them a report. As soon as I know which story has been selected, I send an email to that staffer and to that staffer’s editor and to Jim Witt, the executive editor. I tell them, “Your story has been selected. We need from you a list of the sources in that story and how to contact them so that we can get in touch with them and ask them if they were accurately quoted and paraphrased.”

Then I explain that this exercise is intended to keep our credibility good and strong. I try to put it all in a positive, encouraging context. Because — this has been the case with veteran reporters as well as folks who are on the front end of their career — this kind of thing can be unnerving. I try and keep that in mind and praise them when it’s all through. “So and so gets an A+,” that’s what I’ll put in the subject line.

FG: You try to do one a week?

DH: That’s sort of the ideal. But we can’t really do one a week — especially now, because we just had to cancel our subscription to iThenticate.

FG: Why?

DH: Budget stuff. Belt tightening. Management looks for ways to trim costs and, boy, that was one they felt they could trim. I hated to see it happen, and so did the editor. But we didn’t have any choice. So we’ve had to drop our subscription, which means we’re back to doing it the way we used to do it, which is very slow and tedious. And it’s not as thorough. But it’s still pretty darn good. Depending on the story, it takes us a couple of weeks. Maybe three. Maybe four.

FG: After the whole Jayson Blair fiasco, the New York Times added a public editor position to help key an eye on the paper. Do you think they should add an after-the-fact fact-checking system such as yours?

DH: Sure, to the extent possible. Maybe they only do one story every quarter. But if they have any resources at all to help launch a program like this, it would be time well spent, because it serves as a good reminder internally to staff that the only thing we have is our credibility. Plagiarism is one of the worst things that a journalist can do. I don’t think that is emphasized nearly enough in J-schools, certainly not in high schools. So it’s a way of adding emphasis.

FG: How long do you think the Star-Telegram will keep this up?

DH: It’s a permanent policy. I imagine it will be in effect as long as Jim is the executive editor.

FG: Have you had any pressure from executives at Knight Ridder to wrap it up?

DH: No pressure at all. I think people appreciate the intent and the way it’s done. They agree that it’s a good program.

Felix Gillette writes about the media for The New York Observer.