Hi, I was talking with my colleagues this morning about the news about JAMA retractions.
One article - we can blame the peer reviewers.
Thirteen articles - it's time to blame the editors, and talk about what they should be doing differently.
One thing they can do differently is use professional statisticians, not volunteer reviewers.
We agreed that, for all the money we spend on these journals, it's time we expect the publishers to have their own (paid) in-house statisticians review the submissions as a separate review process from the external peer reviewers. From what we heard from the graduate student interviewed here on CBC, anyone with a B.S. in Stats would have caught a lot of Wansink's malfeasance, p-hacking etc.
What do you all think of this idea?
Melissa Belvadi
Collections Librarian
University of Prince Edward Island