All you scientists frustrated by the rejection of your papers from journals, relax. Rejection before publication is rare, and for those who are forced to revise and resubmit, the process will boost your citation record. At least, those are the results of a massive survey of scientific authors published online yesterday in Science.
The idea for the survey came to Vincent Calcagno, an ecologist at the Institut Sophia Agrobiotech in Sophia-Antipolis, France, in 2008. At the time, he was a postdoc at McGill University in Montreal, Canada, and experiencing "the frustration of having a piece of research I liked a lot rejected by several journals (six or seven) in a row." Calcagno heard similar horror stories from his colleagues and wondered "if this was frequent, if this was worth it, if research was losing energy in such long publication processes." He poked around for data and discovered that there was none. Everything that happens to a scientific manuscript before publication is "a black box," Calcagno says. So he decided to create those data himself.
Calcagno wrote a computer program to harvest the e-mail addresses of the more than 200,000 corresponding authors on papers published between 2006 and 2008 in 923 journals, all of them in the biological sciences or multidisciplinary journals such as Science, Nature, and the Proceedings of the National Academy of Sciences. The program sent each author an e-mail asking whether the paper had been submitted previously to a different journal, and if so, the name of that journal.
The replies came in such a torrent that McGill's e-mail servers temporarily shut down. About half of the authors responded, and after excluding replies that his programs could not decipher, Calcagno had a data set covering the prepublication history of 80,748 biological science papers. To analyze the data, he teamed up with researchers at McGill and elsewhere.
The study's basic assumption, Calcagno says, is that scientific publishing can be studied as a network, much like the ecological food webs that he studies. Journals, for instance, behave like predators competing for the common food resource of scientific manuscripts.
Some of the interactions that the researchers observed confirm common assumptions about scientific publishing. For instance, the more prestigious journals, as measured by their impact factor (IF), get the lion's share of first submissions. And after being rejected by one journal, authors tend to next submit to a journal with a lower IF.
But there were some surprises. For example, the painful experience of endless rejection before publication that motivated Calcagno in the first place turns out to be rare. About 75% of published papers were accepted by the first journal to which they were submitted. Even more surprising was what happened to rejected papers once they were accepted elsewhere: The previously rejected papers had a slight bump in the number of times they were cited by other papers. Calcagno speculates that this is the effect of the review and editorial process. When scientists are forced to rewrite and answer tough questions, the quality of their papers improves. The practice of resubmitting a paper rejected by one high-profile journal, such as Science, to another, such as Nature, may also play a role, since it increases the odds that the paper will still end up in a relatively high-profile journal. (Ironically, Calcagno says his new paper won't get this advantage, since it was submitted to only Science.)
Correction, 15 October: A previous version of this article stated that "rejection is rare." Because the study focused only on already published papers, rather than all papers submitted to journals, the more precise statement of the study's finding is that rejection is rare before publication. The article has been changed to reflect that.