Are you done hunting scientists for the day?
I haven’t had time today. But if I spent a couple of hours reading papers, I’d probably write four or five comments about errors in scientific papers. It’s not hard to find these problems, and it’s not hard to find them at any institution. They’re all out there to find, if anyone wants to read the scientific literature.
What motivates you to spend the time to do it?
I’m not an anti-vaxxer, I’m not a crank conspiracist or anything like that. I’m a scientist myself, and I care about getting the science right.
You focus in part on images that have been mislabeled, or used twice in a paper but as evidence for different things, lazily cut-and-pasted from one spot to another. Is there a simple way for you to tell that an image is wrong?
You just look at the picture and read the labels. For example, if you look at a microscopic picture of cells, you see the position, location, orientation and shape of the cells. And if you look at another picture of cells and they’re all in the same position, with the same shape and orientation, then you know that this is the same image, right? It’s not a complicated process.
You’ve also identified errors in western blots. What are those?
These are a type of scientific experiment used to identify and quantify specific proteins. The images are important in a lot of scientific papers. They look gray in the background and have black bands. When you look at them very closely, you can usually tell whether it’s a copy-and-paste job or not. These things aren’t always obvious to people who don’t look at a lot of western blots.
Let’s turn to Dana-Farber. After finding errors in multiple papers from its researchers, what inference do you draw about the scientific methods of that pre-eminent institution?
It’s important to remember that Dana-Farber researchers publish a lot of papers. But it’s still a lot of errors, and they’ve happened over a long period of time. This tells me that for a long time people haven’t paid close enough attention to getting the basics right. How many sloppy errors are we comfortable with top institutions making? It’s probably not many. I think most people expect that Harvard scientists aren’t doing copy-and-paste mistakes often.