Via a comment by Christina Pikas, there’s a post at the Scholarly Kitchen about a new study quantifying the use of the arxiv:
Employing a summer intern, Ingoldsby conducted an arXiv search of nearly 5,000 journal articles published by the American Institute of Physics and the American Physical Society. Their methodology was painstakingly robust, looking for title variations and having all unsuccessful searches repeated by a trained physicist.
The percentage of articles found for each journal in their studied varied greatly. While fields such as elementary particle physics and astrophysics reported nearly 100% overlap, this finding was not generalized over other sub-disciplines in physics. Many fields showed much less coverage, many under 5%.
I go back and forth in what I think about the arxiv. It’s certainly a useful tool, though far from indispensable in my areas of interest. I’m also a little dubious about its effect on the culture of academia. On the other hand, it is the primary vehicle for dissemination and discussion of papers in high energy physics, which is the field with the best representation in blogdom, so I’ve argued several times that ResearchBlogging.org ought to allow tagging of posts based on arxiv preprints.
This article is a nice reminder to keep things in perspective. The arxiv is dominant in some sub-fields of physics, but hardly used in others, and by no means does it even provide a representative sample of the current state of physics. It’s getting better all the time, but there’s still a long way to go.
Oh, and check out Christina’s blog, if you haven’t already.