Projects with bigger potential impact are more likely to be published in CNS or as headlines - but are also much riskier and more challenging. Not a surprise that more of these would turn out to need followup. That's why we need to have standards for reproducibility. - Shirley Wu

I think there are 2 wrongs here: one as in the results are incorrect, the other as in so hyped up by the media report that the conclusion reached by readers will be wrong. - Neil Saunders

+1 Neil. Media creates unrealistic and sometimes faulty expectations. Maybe science blogging can ameliorate that!? - Shirley Wu

Where is the evidence that "the results are incorrect"? Not in this Economist piece. And not in the PLOS Medicine piece on which it is based. (That refers to clinical trials papers, but not to the scientific literature.) So where is the evidence for what you are writing in the comments above, or is it opinion? ;-) - Maxine

It's been a bugbear of mine for a long time. Let's forget the wrong part. Media tends to either go for doom and gloom, or complete overhyping - Deepak Singh

Ah, media, well I am not going to argue about that! But the link here is about the scientific literature specifically. As it is a media article, though, to quote Deepak, it must be going for "doom and gloom, or complete overhyping" ;-) - Maxine

Just to clarify: I was referring to an abstract notion of "results being incorrect" in order to compare with the over-hype issue, not to this article or any one publication in particular. - Neil Saunders