Links. Photos.
Fraud. Fraud. Fraud everywhere I look
Links of the week
Is it Acceptable to Lie for a Good Cause? Some of these examples of lies are claims I had seen many times in the wild and it’s interesting to see where they come from.
This post on scientific fraud originated a lot of discussion on Twitter. There is at least one counter-claim (from someone close to the original events) that the piece simplifies the story to make the fraud more blatant than it was (and make the PI of the story look negligent), but there was also a sprawling discussion on how much responsibility the PI (senior author on the paper) bears in this case.
My opinion is that PIs bear responsibility1 to implement at least some measures to protect themselves (and the scientific community at large) from this type of event. There is a lot of gnashing of teeth about how this would mean that labs turn into places of little trust, but I frankly do not see much of a tension because the exact same practices that catch fraud are also good for catching mistakes and making science reproducible. For example, internal sharing of results and making sure that multiple people have a look at the data and analysis makes the science stronger in the best of cases and catch fraud in the worst case (here’s a report that the fraudulent postdoc was resistant to working with others in this case),2… Even if you are not worried about fraud per se, these are all things that you should be doing to catch honest mistakes and they will catch 100 honest mistakes for each one instance of fraud that they catch (because honest mistakes are so much more common).
In this story, it is actually the other trainees of the same PI that seem to have had their careers disrupted the most (including a lot of self-doubt when they could not reproduce the fraudulent results).
Re-posts of the week
Since, it’s Nobel Week, here is my yearly repost of “Why BLAST should get a Nobel Prize.”
Related to the link above, I had already posted about the risk of fraudogenic scientific cultures
Photos of the week
From a trip to New South Wales:
This does not mean that the PI has strict liability whereby they are automatically responsible for any fraud that occurs in their lab. But they should be ready to answer questions about how did this happen on your watch?
One interesting practice in finance is mandatory holidays with (project handover). Apparently, people who never take a holiday are often hiding something and handing over your projects for a few weeks makes it easier to catch fraudsters.