< Manipulating Science Reporting


Friday, September 28, 2012

BROOKE GLADSTONE:  Last week you may have come across some news reports about French scientists finding a link between genetically modified food and cancer. Their research was announced at a press conference on the very same day it was published in the Journal of Food and Chemical Toxicology. That’s not unusual. What is unusual is that in order to get an early peek at the study, reporters had to sign a nondisclosure agreement that made it impossible for them to get other experts to weigh in on the research. This manipulative practice, says science writer Carl Zimmer, ensures that the scientists can, quote, “bask” in the badly reported media spotlight. He says that the French study, which involved feeding genetically modified corn to rats, was far from perfect.

CARL ZIMMER:  So the control rats, 30% of the males and 20% of the females, developed tumors. The rats that got fed the genetically-modified corn, 50% of the males developed cancer and 70% of the females developed cancer.

BROOKE GLADSTONE:  Well, that’s a lot more cancer.

CARL ZIMMER:  Sure, it sounds like a lot more.

BROOKE GLADSTONE:  It is a lot more, right?

CARL ZIMMER:  Well, is it a lot more, statistically speaking? Is it so much more that you can start to say definitively that genetically-modified food causes cancer? And, actually, you can’t, in this case. In fact, the paper does not actually show whether or not the cancer rates are statistically significant. That’s something that you just expect to see in any paper like this.

BROOKE GLADSTONE:  The French study was published in a peer review journal. Doesn’t that ensure that scientists who aren’t affiliated with the study have already had a look at the research and deemed it credible?

CARL ZIMMER:  Peer review is important but it’s no guarantee that the science is right. Science reporters need to talk to other scientists whenever they’re reporting on research. And it’s especially important when it’s extremely controversial.

We’ve had these huge debates about genetically-modified food, screaming matches around the world. In California, actually, they’re considering whether to label genetically-modified foods, and people who are advocating for that immediately grabbed onto this study. As soon as this study was in the press, the French Prime Minister said that if it was confirmed, that France would push to ban genetically-modified corn in all of Europe. So you have to be able to check with other scientists, other experts to see whether the science really holds up. And in this case, as soon as other scientists took a look at this paper, they said, whoa, this is a mess.

BROOKE GLADSTONE:  Why was it a mess?

CARL ZIMMER:  The groups of rats that were being given different treatments were very small. There were only ten rats in each group. The rats that they chose are a special breed of rats that’s actually very prone to getting cancer [LAUGHS]. This is actually important for scientists who want to study cancer and how tumors develop, so these are an important breed of rats. The problem is that if you’re only using these small groups of rats, you might just end up with one group of rats that have a lot of tumors just because that’s how the rats are. This kind of study is often done with many more rats in each group, say like 50 or even 100. It’s expensive, it’s difficult but if you want to get the statistics right, you need more rats.

BROOKE GLADSTONE:  What really set the fire under you is not that journalists were given this article in advance, but that they had to sign a confidentiality agreement. In other words, this was a rare example of not being able to do any advance reporting on a scientific study.

CARL ZIMMER:  Right. The only advance reporting that people could do was just to listen to what the authors of the study told them. The paper was embargoed until the middle of our press conference, where they were announcing the results. That way they were ensuring lots and lots of coverage skewed their way. It would be sensationalized because there wouldn’t be any way of talking to other scientists. There’s been follow-up reporting that has focused on the controversy but it’s really important to make sure that the first reporting on something is done right because a lot of times that’s just what people look at.

BROOKE GLADSTONE:  And you suggest that journalists just walk away, if asked to sign a confidentiality agreement.
CARL ZIMMER:  Absolutely. This is wrong. I learned that the BBC had done exactly this. They had been offered the paper a day in advance, if they signed this confidentiality agreement, and they just said, forget it and they walked away.

BROOKE GLADSTONE:  But if you walk away, you also don’t have any opportunity to offer scientific opinions because you don’t have  enough preparation time; you just miss the cycle.

CARL ZIMMER:  Well, you’re not gonna be doing any legitimate reporting if you take that paper and can’t talk to anybody else. What you’re gonna come out with is bad journalism. And what journalist wants to come out with bad journalism?

BROOKE GLADSTONE:  It’s about the hits!

CARL ZIMMER:  [LAUGHS] Well, I suppose that there are reporters out there who get pressure from their editors to cave in and just sign these things so that you can get the hits right away. But those aren’t places worth working for.

BROOKE GLADSTONE:  Carl, thank you very much.

CARL ZIMMER:  Thank you.

BROOKE GLADSTONE:  Carl Zimmer teaches science writing at Yale University and writes The Loom blog for discovermagaine.com.


Carl Zimmer

Hosted by:

Brooke Gladstone