Last year I introduced the concept of desearch, which sadly but predictably has failed to take the world by storm. Desearch occurs when someone produces research so bad it actually weakens the case they claim it is supporting.
In the last few weeks Dr Judy Carman has published two papers, one as lead author. Carman has a background in epidemiology and is both an opponent of the use of GMOs and a researcher exploring their drawbacks.
Some might leap to the conclusion that these two things are in contradiction, and that Carman’s credibility is undermined by her activism. I disagree. If a researcher comes across preliminary evidence for something being dangerous it is entirely right and proper for them to both explore the evidence in more detail, and campaign for a moratorium on the source of the danger while the research is done. So no criticism from me there, but it does mean any results she comes up with that are bad for GM deserve closer scrutiny.
The thing is this: Both the papers Carman published look, on face value, very bad for GM. If either of them prove to be right GM foods, at least as currently used, are in a lot of trouble. But one of these papers looks to be loaded with flaws. Deeper scrutiny suggests it proves nothing at all, and barely even suggests a problem. The other paper is, if anything, looking stronger with time. The only post I can find attempting to rebut it fails to come to grips with its argument at all, suggesting if there is any weakness it is hidden from leading GM supporters.
So my question is this: Does Carman’s authorship – lead authorship at that – of a highly unconvincing paper discredit an apparently much better paper on which she is co-author? I’m not sure and will ponder further.
Meanwhile, a little on the two papers. I’ve blogged about the one paper before. It does not explore the safety of GM at all, instead looking at whether the GM products on the market have produced much value. The answer seems to be no. While yields have risen since GM crops were introduced in the US and Canada, they remain lower than in non-GM Europe for canola, while Europe has actually roughly caught North America in corn yields. GM doesn’t seem to be the cause of the problem, since Europe was improving relative to North America before GM came along, but it has done little if anything to help, while leading to vastly greater use of herbicides with potentially serious side effects.
The other paper is this one, suggesting that GM foods cause stomach inflammation in pigs. While aspects of the trial were well done, with proper control groups etc, there are a lot of problems. Perhaps the most noticeable one is that the inflammation was divided into four classes. The GM pigs had more of the highest level of inflammation, but were also more likely to have none at all. While I don’t think the snark about GM opponents saying GM has no nutrients is justified, this post otherwise does a much better job than I can of explaining why the study really proves nothing, complete with excellent XKCD contribution.
If Carman really thinks she is on a winner with this study her judgement becomes quite suspect, and that raises questions about her other research. Which is a great pity, because Carman has been doing a good job of trying to conduct double blind controlled feeding studies of GM foods. This is something I think we need more of, and not many people, for a wide variety of reasons, are stepping up to do.
The other thing to note here is that GM opponents have jumped on the feeding study and circulated it widely, while the yield paper seems to have attracted much less attention. A hint – if you want to discredit GM, try publicising the paper that actually does that.