My post last week about citation rates of mandated vs. self-selected Open Access (https://web.archive.org/web/20101124191039/http://blogs.plos.org/mfenner/2010/10/18/new-in-plos-one-citation-rates-of-self-selected-vs-mandated-open-access/?) resulted in an interesting discussion thanks to some good arguments made by Stevan Harnad. One personal conclusion for me: mandates for self-archiving are not a good idea. I would very much prefer researchers to be highly motivated to self-archive thanks to a repository that is both fulfilling important functions and is fun to use.
What follows is a short list of ideas – many of them obvious and some of them already implemented – that for me would make repositories more attractive to use.
Disciplinary and institutional repositories are good places to make primary research data publicly available. Particularly if no standard database exists, such as GenBank (https://web.archive.org/web/20101124191039/http://www.ncbi.nlm.nih.gov/genbank/?) for genetic sequence data. Whereas everybody nowadays talks about making research data available (https://web.archive.org/web/20101124191039/http://www.jisc.ac.uk/whatwedo/programmes/~/link.aspx?_id=28A2778937C74EB285F05E38BFBD5DEE&_z=z?), the threshold to do so is often too high for more specialized datasets. An institutional repository is a good place for simple grassroots solution.
PubMed Central (https://web.archive.org/web/20101124191039/http://www.ncbi.nlm.nih.gov/pmc/?) and UK PubMed Central (https://web.archive.org/web/20101124191039/http://ukpmc.ac.uk/?) are very popular disciplinary repositories for the biomedical and life sciences. Unfortunately I can only submit manuscripts to them if the research was funded by one of a few funders (including the NIH, but none of the major German funders).
Deposition of manuscripts prior to peer review has a long tradition (https://web.archive.org/web/20101124191039/http://blogs.plos.org/mfenner/2010/10/09/new-on-arxiv-first-report-of-soap-open-access-publishing-project/?) in high-energy physics and related disciplines. Preprint archives probably don’t work in all disciplines, but I have argued before (https://web.archive.org/web/20101124191039/http://blogs.plos.org/mfenner/2010/10/16/in-which-i-suggest-a-preprint-archive-for-clinical-trials/?) that they could be a very good idea for clinical trials.
While most publishers allow some form of self-archiving by authors (pre-print, post-print and/or publisher’s PDF version), there are unfortunately still exceptions. The American Chemical Society (ACS) has a particularly unwelcoming policy (https://web.archive.org/web/20101124191039/http://scientopia.org/blogs/bookoftrogool/2010/10/05/acs-the-perfect-storm/?).
An institutional repository can host pre-prints of submitted papers. But the repository could also enable a tighter integration with the journal submission process. The German eSciDoc (https://web.archive.org/web/20101124191039/https://www.escidoc.org/JSPWiki/en/Overview?) project uses that approach. Most researchers would probably welcome technical and financial assistance in the submission process.
The institutional bibliography showcases the scholarly work done by a particular researcher, research group, department or institution. Many researchers like to see up-to-date profiles (including publication lists) on their institutional webpages. Bibliographies are also collected for evaluation purposes. We have started to use BibApp (https://web.archive.org/web/20101124191039/http://www.bibapp.org/?) at our institution. Repositories can facilitate fulltext access to PhD theses and other publications in the bibliography.
Regular readers of this blog know about my involvement (https://web.archive.org/web/20101124191039/http://blogs.plos.org/mfenner/2010/09/07/orcid-as-unique-author-identifier-what-is-it-good-for-and-should-we-worry-or-be-happy/?) in the Open Researcher & Contributor ID (ORCID) initiative. All scholarly contributions, including repository content, should be unambiguously connected to their creators as this will greatly facilitate the discovery process. I look forward to ORCID support in repository software such as DSpace (https://web.archive.org/web/20101124191039/http://www.dspace.org/?) and EPrints (https://web.archive.org/web/20101124191039/http://www.eprints.org/?).