I have come to appreciate the important role that prestigious journals like Science, Nature and Cell play in filtering out bad science, and protecting both the public and other researchers from wasting their time reading about – or following up on – results that are not believable. You have all, undoubtedly heard about recent studies examining the reproducibility of scientific results. For example, a recent Nature paper [paywalled, so you can believe it] described how scientists at drug company Amgen were able to successfully replicate six of 53 landmark studies in cancer research. As these were landmark studies, most were published in the highest profile subscription journals. And these results prove that – contrary to what I would have expected – the top subscription journals doing a great job of picking papers. First, Amgen, who doesn’t like to waste their money, found 53 of these studies important enough to try to replicate. I don’t think they’ve bothered to try even a dozen PLOS ONE papers. But more amazingly these scientists at Amgen were able to get the same results as important academic scientists OVER ten percent of the time. This means that the papers must have described the methods extremely clearly – a hallmark of high profile journals.

Why I, a founder of PLOS, am forsaking open access Thursday, April 3, 2014 @ 11:11am