Nature and Science know best

Abstract

For many researchers, getting work published in Nature or Science is akin to finding the Holy Grail. Aside from acquiring some high-profile international recognition for one's research, having a paper in one of these journals on your CV, particularly early in a career, can greatly (some might say disproportionately) improve your chances of future employment, promotion and funding. Employers and funding bodies might assume that individuals with Nature or Science papers are more likely to produce substantial bodies of internationally regarded work in future. Those of us who are less fortunate often wonder whether this assumption is justified.To examine this issue, I perused departmental websites and the Web of Science (http://wos.mimas.ac.uk/) and examined the track records of academic and research staff from the life sciences departments of British and Australian universities who began their careers (i.e. started publishing) in the early 1990s (1990–1994). I focused specifically on workers in the fields of ecology and evolutionary biology, partly through personal familiarity with these subjects, but also because workers in other fields, such as medical research, have several journals of equivalent impact in which to publish original research (e.g. Cell, New England Journal of Medicine, etc.). Although ecologists and evolutionary biologists can also publish in other high-impact journals, such as TREE, these tend to be review journals. The researchers were divided into those who did, and those who did not publish in either Nature or Science from 1993 to 1998. I then counted the number of publications in scientific journals that the researchers had produced since 1999. I also noted how many of these papers were published in Nature or Science.The differences were striking. Researchers who published in Nature and/or Science early in their careers (38 out of 259 included in this survey) were, on average, almost twice as productive in the past five years compared with those who did not (mean numbers of publications=21.08 and 10.99, respectively; one-tailed unequal variances t-test: t=−4.27, df=40.471, P just over two thirds of researchers in this sample who have yet to publish in Nature or Science) can at least console ourselves with the fact that we are in the majority.