Agar slants in tubes labeled with Sharpie marker.
Conservatively, about 2% of the published scientific research data is fake, at any one snapshot in time, if I’m reading the report by Fanelli et al. correctly. And that’s just for the researchers who didn’t want to look like they were lying in response to a questionnaire, “Do you ever lie?” and most said, “Um, sometimes.” (The study was a retrospective analysis of a bunch of previous surveys, but still, it was self reporting — and if you say, “No, I never lie,” you’d look like you were lying so who knows.). Ok. I get that. I’ve seen that.
(Your loyal bloggist gave up being a scientist because I was an unqualified disaster in the lab, plus, the start up companies I worked for at the time bounced my paychecks. So I never faked data, it wouldn’t have helped anyway. Once I injected myself with something by accident. To this day, I still look between my knuckles to see if there’s any weirdness. I can’t really tell. Every time I come down with something, in the back of my mind is “I wonder if it was that supernatant that I injected a long time ago?”. So my failed scientific career is a monkey on my back. Probably the worst was when I stayed up all night growing some cultures and spinning out some DNA, and I had an ethanol precipitation in the microfuge going on and one of the tubes opened up and washed away all the labels (I must not have used a Sharpie which only washes off with acetone, that I now in my post-science life refer to as nail polish remover.) It wasn’t the staying up all night that bothered me it was having to tell my boss that I screwed up the labels and didn’t know what was in what tube. True dat. Moreover, the act of putting a label on the reagent bottles was something of a fine art in that lab, as our student intern decided to label the 7N NaCl “Earle Klugh” and the KI was “George Winston” and tween or some detergent was “Michael Jackson” (back before he fell from grace). Sigh. )
But. I think science fraud is more prevalent than 2-3%, mostly because of people’s reactions to me. I may not be a scientist, but I’m not overly stupid, either (usually). And some scientists tend to be arrogant and lay on the jargon specifically to set up a barrier to entry for people like me. If someone can’t explain to me what they’re working on in a few minutes so that I understand it, they’re usually trying to buffalo me. It’s not very often, but every once in a while. . .
Now, I’ve seen some doozies. Usually, there’s a Dr. Frankenstein who bullies the Igor into doing the dirty work, and sometimes it’s not even fake data, sometimes, it’s just lying about what results mean. Like, when you claim 99.999% purity, but that’s with respect to your starting source. Where you look at what’s in the test tube, you have 50% pure, 50% junk. I won’t go into what name brand editorial boards of name brand peer reviewed publications can do if you submit something that maybe one of them has file for a patent on. . .
But I don’t think the scientific system is any more, or less, free from fraud and greed than the financial industry or any other industry. It’s just harder to understand and there tends to be layers of secrecy.
Actually, having just thought about that last sentence after I typed it, maybe that’s not even true.
The sad thing is that usually, they don’t need to fake any data. Usually they’re on to something — it’s just not as unambiguous as they would like. Everyone likes nice clean photos of a fat, smiling, brightly-lit cell saying “See? Here I am. You’re a genius.”
Usually, however, there’s all sorts of background goop, and the cell is not smiling, and it’s looking dull and hungover and saying “whew where am I? Who are you?”. There is a temptation to just blot out non-specific binding.
But even if you do blot out the background to make your result seem more clear, it doesn’t mean that there isn’t a result. Leave the background in there. It’s a methods problem. Solve the methods problem and you get two papers or patents, instead of just one fake one. See?
My sense is that the really outre fakers probably are pretty rare, and more common are the fakers who enhance results to bring them into high relief. Like being in the top 3 for the Tour de France and taking EPO. You’re almost there, and you just want to get some more oxygen to make sure.
Comments