This is a commentary on Henk Moed (2012) "Does open access publishing increase citation or download rates?<http://www.researchtrends.com/issue28-may-2012/does-open-access-publishing-increase-citation-or-download-rates/#comment-41>" *Research Trends* 28
PHYSICS, THOMPSON-REUTERS ISI AND DOWNLOAD ANALYSIS All the points made in Henk Moed's overview of the effect of open access (not just "open access *publishing*!) on citations and download below are very welcome, timely and valid. Just three complementary comments: 1. PHYSICS. Physics is a field with very high, un-mandated self-archiving rates for over 20 years (perhaps as high as 80-90%), compared to other fields (about 20%). Physics (and astronomy) are also fields with relatively high journal accessibility levels, compared to other disciplines. So the explanation of the particularly high citation advantage in physics and astrophysics is very probably due to the fact that in those fields most papers are being made OA un-mandated and only the weakest papers are (self-selectively) not being made OA. Hence the OA/non-OA difference may reflect a large element of strong/weak research difference, rather than just an OA accessibility advantage, leaving earlier OA (for pre-publication preprints) as the only OA factor in the difference. In contrast, in most other fields there is more journal inaccessibility (because of subscription un-affordability) and much less OA self-archiving. Hence the OA citation advantage, though not as big as in physics, is always positive in every field, and equally great when mandated or un-mandated (self-selective). 2. THOMPSON-REUTERS ISI. To the extent that the OA citation advantage is based on Thompson-Reuters ISI-indexed journals, it is indeed based on the top journals and the journals to which researchers are most likely to have access. Hence it may well under-estimate the size of the OA advantage. Some of the studies in the OPCIT bibliography cited by Henk were based on SCOPUS and even on Google Scholar, but there is certainly more scope for a broader analysis of unindexed journals too. (I would disagree with Henk, though, that the access of potential users to even the ISI journals is anywhere near good enough. The fact that we find a significant OA advantage even with just ISI journals would seem to confirm this. The likely cause of the enhanced OA citations is the enhanced access provided by OA. But looking directly at the relation between journal affordability and subscribership in the OA advantage is certainly a good idea.) 3. DOWNLOAD ANALYSIS. Henk is quite right about the need for more usage analysis in connection with OA. As OA grows (with the adoption of more and more funder and institutional self-archiving mandates) institutions will have both the record of what their researchers publish annually and when and where it is being downloaded. However, there is a fundamental difference between comparing OA/non-OA citations and OA/non-OA usage: Citations are citations, no matter where they are measured. But usage is locus-specific, depending on whether the locus is the publisher's website, the institution's online subscription database or the repository of the institution's own output. Institutional repository downloads are hence the sound of one hand clapping, since, by definition, those papers are all OA!
_______________________________________________ GOAL mailing list GOAL@eprints.org http://mailman.ecs.soton.ac.uk/mailman/listinfo/goal