Astronomers lead the pack on Arxiv June 18, 2009Posted by Sarah in astro 2.0, science.
Tags: arxiv, astronomy, dotastronomy, open science, publishing
How much of the research record in physics can actually be found on Arxiv? How accurately does Arxiv reflect what is happening in physics today? On the whole, not so well, says the American Institute of Physics. Over at the Scholarly Kitchen, Philip Davis reports on a presentation given at a science editors’ conference by Tim Ingoldsby of the AIP that shows a highly inconsistent coverage of the literature on Arxiv between the different fields of physics research.
For the influential journal Physical Review Letters the results for astronomy were encouraging, with over 90% of all journal articles being available via Arxiv. This was bettered only by the “Elementary particles & fields” category, for which the reported coverage was 100%. In contrast, in the hugely dynamic and competitive field of Condensed Matter Physics less than half of all PRL papers were found in the Arxiv. For the category “Plasma & beams” the number was a measly 10%. I wonder why there are such big discrepancies? What causes this difference in culture between different research areas?
I’m not sure of the number of astronomy papers involved, as PRL is not exactly an obvious journal to publish in for astronomers. So these papers are may not be a representative sample of all astronomy literature. However even a careful suggestion that astronomers are leading the way in physics in embracing a culture of openness in research I think is a great development.
[Update: I exhanged a couple of emails with Tim Ingoldsby and he was kind enough to send me an overview of the data he used in the analysis. The number of gravitation and astronomy papers in the 12 issues of PRL under study was 22 – just 2.7% of the total content being studied. As expected, it’s a small number, but not totally insignificant.]
Much of the interest in Arxiv comes of course from the publishing industry, who believe that the fast-expanding repository (see figure) hurts their industry. After all, in a world where > 90% of all physics research is available gratis and for free via Arxiv, who wants to pay an expensive journal subscription? Although I think the traditional way in which we report our results to the community is rapidly becoming outdated, with long publication times and imperfect peer review, Arxiv in its current format can’t and shouldn’t replace this. Peer review is a crucial part reporting results, and in the majority of cases results in a better paper. Indeed, Inglesby’s study showed the incompleteness of the Arxiv record, with authors not going back to update their papers once they’re definitively accepted by the journals. But I would love to see more ideas on how to innovate this business.
This is not something the journals can do single-handedly though – they do try (see e.g. here for some examples). I think the science community itself is conservative about adopting new technologies for talking about science. And that in turn has to do with the way science is funded and how we measure success in research. Physicist Michael Nielsen wrote an excellent article on this in a recent issue of Physics World, and Rob discussed the piece at length in a new blog post. Incidentally, Nielsen is an invited speaker at .Astronomy, and I’m looking forward to discussing these issues at the workshop.
Ingoldsby’s presentation is available online here.