“Salami slicing” refers to the practice of breaking scientific studies down into small chunks and publishing each part as a seperate paper.
Given that scientists are judged in large part by the number of peer-reviewed papers they produce, it’s easy to understand the temptation to engage in salami publication. It’s officialy discouraged, but it’s still very common to see researchers writing perhaps 3 or 4 papers based on a single project that could, realistically, have been one big paper.
But I’ve just come across a salami that’s been sliced up so thinly that it’s just absurd. The journal Archives of Iranian Medicine just published a set of 33 papers about one study. Here they are – this is a rather silly image, but it’s a silly situation:
Yes, a single survey of the mental health of the Iranian population has been published 33 times. 31 of these papers (all published on the same day) are devoted to the Iranian provinces – there are 31 provinces of Iran, and each province got its own paper listing the results from that area. The author lists are more or less identical every time.
This degree of slicing is not standard practice when publishing national epidemiological statistics, for obvious reasons. For one thing, it will make it difficult for readers to compare rates of disease across different provinces. They’d need to consult all 31 papers. Then they would need 31 citations to refer to all of those papers if they publish the comparison. It’s just not practical.
Writing the 31 provincal papers would also have been a huge amount of work for the authors, although they seem to have cut some corners here, because the papers contain a lot of overlapping text. I guess however you slice a sausage, all the pieces contain the same meat. Here’s a sample of the overlap:
As well as the 31 province papers, there are 2 other Archives of Iranian Medicine papers that sum up the whole mental health survey (1, 2). One or two summary papers like this are what you’d expect to see from a study of this kind. The provincal data would usually be included in these papers as a supplementary material.
Oh, and one of the summary papers has already been cited 34 times – guess where 33 of those citations come from?
So why did Archives of Iranian Medicine allow this? On Twitter, it was spotted that one of the authors is an associate editor at the journal, which may be relevant. The Editor-in-Chief of Archives of Iranian Medicine is not listed as an author, but he is thanked in all of the papers for his “comprehensive support” of the project.
Overall, this is the worst case of salami slicing – or sosis slicing – I’ve ever seen.