Open Access News

News from the open access movement


Tuesday, May 02, 2006

Tips for filling an OA repository

Arthur Sale has distilled his experience as a researcher, founder/manager of an OA repository, and student of repository policies, into an excellent list of suggestions for getting authors to deposit their work. Excerpt:
Effective author support policies involve a plethora of activities, and are well exemplified by the activities undertaken at QUT, Queensland University and here [University of Tasmania]. No doubt in many other places. They include (but no university does all):
  • Assistance with uploading the first document (hand-holding). Maybe devolve this out to departments/faculties/workshops.
  • Fall-back positions which allow a subject-librarian, or a department/faculty office professional, to upload on behalf of an author who is not computer literate.
  • Provision for turning final manuscripts into pdf format (info about free OSS options and/or a library service).
  • Provision of as much (automated) statistical use information as authors find useful. See [this] for example.
  • League tables of document downloads (Do NOT publish or put on the Web league tables of academics by totals of downloads. This is counter-productive as the same few people are always at the top {sometimes because of extraneous discipline or popularity reasons}, and everyone else feels aggrieved). Document download info seems ok as it is anonymized and variable. See [this] for example.
  • Encouragement (or stronger) from a head of school or research coordinator - they need to be converted and they are intra-university competitive as well as being discipline-competitive.
  • Integration of the repository into school and university websites (eg instead of a list of publications on a web-page (always out of date) put a php/perl query on the repository for the particular author or authors (always up to date). Possibility needs promotion and education to web-page designers (may be academics).
  • Professional development workshops for PhD candidates to put their publications up (Important: these are Trojan horses. Maybe you can get a mandate for them ahead of academics/faculty)
  • Development of repository software to provide extra information to authors and possibly readers, such as citation counts.
  • Briefing meetings with heads of departments, deans and research directors. Keep it as routinized as possible: we are not trying to do something radical but to smooth something that should be a routine part of research activity.
  • When you have a mandated policy, act on selected departments/faculties in a sequential strategy. Do not attempt a scattergun approach. Again, it is routinization that you are after.
  • Some universities have introduced financial benefits for depositing.
  • Do not worry about metadata quality, nor bother authors about it. Authors are often as good as librarians, if not better. In any case the most popular discovery techniques are not dependent on metadata.
  • Provide a service for authors who are worried about copyright. It generally isn't important nor is the service onerous....

Getting back to the requirement (mandatory) policy. I well understand that most universities do not yet have such a policy. I think I know exactly how many do. However, unless it is in your kitbag (like a field-marshal's baton) the university is wasting its money even having a 5-15% full repository. Striving to achieve such a policy is understandable and laudable, but it must be a continuous and strong push.

However, expending money on author support policies without a mandate is like pushing a large rock up a hill. It does not work and is demonstrated not to work. Precisely because of what I wrote earlier: the vast majority of academics (85%+) are non-participants and will seize any excuse however spurious to avoid doing any extra work. They are incapable of being persuaded in the mass. Remember that I am a researcher, not a librarian. I know the mindset of researchers. So to summarize:

  • Try to get the mandate before the repository.
  • If you've got the repository before the mandate, make it crystal-clear to everyone (especially in higher management) that a mandate is in your sights and you are not going to let go of it until you get what you want and the forces of reaction are defeated. Use the word "luddite" if you have to.
  • Don't expend significant amounts of time and money on author-support until you've got the mandate. It is pretty much wasted anyway, like flushing dollar notes down the toilet.
  • After you've got the mandate, go for full-on author-support. It will speed up the transition which will take 1-3 years.