Digital Repositories and Preservation

Jingfeng Xia, A Comparison of Subject and Institutional Repositories in Self-archiving Practices
The Journal of Academic Librarianship
Volume 34, Issue 6, November 2008, Pages 489-495

In this article, Xia reports the results of a small investigation into the self-archiving practices of physicists at the University of Southampton. He specifically investigates contributions made by researchers to the arXiv subject repository vs. contributions made to Soton, the e-Prints-based institutional repository at Southampton, with the hypothesis that active contribution to an SR will be mirrored with active contribution to a local IR. While the results are fairly interesting, I specifically found some of the observations made in the article worth mentioning and discussing.

In the introduction, Xia mentions the few mandates that have been created, specifically NIH and Harvard University, but states “it still remains questionable about how such mandate policies can be implemented in practice,” which I feel is a thoughtful statement. Much like IRs have not been successful merely by their creation, a mandate policy is not a magic bullet without successful implementation - likely requiring quite a bit of effort by a number of individuals within an institution. If a mandate was somehow created and implemented here at Mason, I can’t imagine how I would even begin to manage the amount of data that would be deposited and ensure it’s legality, especially with the nearly complete lack of support I currently have (and I’m hardly the only IR manager who would say this).

Xia makes some of the points that bear repeating: self-archiving hasn’t brought about a satisfactory number of content materials to the vast majority of IRs (although the definition of “satisfactory” varies between institutions), and faculty do not show interest in self-archiving, even with the increased awareness of IRs. Additionally, Xia mentions something that I think most upper administration don’t think about: “Among the operational tactics of the IRs that have played a crucial role in the making of the content are a liaison system and a mandate policy” [bolding mine]. This is becoming so blatantly obvious by the research and reports by managers, that I can really no longer pretend that my solo operation will keep the IR here limping along. I need the assistance of my fellow librarians, but that requires a fairly major shift in the way the IR is viewed by the University Librarian on down. Liaisons aren’t necessarily going to place a high value on scholarly communication, and may feel too busy to deal with gathering content from researchers or managing deposits. This is compounded with the frustrating (yet necessary) copyright issues that frequently require a substantial time spent on Google or waiting on a return phone call or email from a little-known publisher. I don’t fault liaisons for not jumping in to help with the IR, but it’s becoming obvious most IR managers need their help.

Xia’s method of comparing deposits to the SD vs. the local IR were quite thorough, and he ensured to check against name variations and verifying the affiliations of authors and such. His findings, while perhaps unsurprising, bear discussion. As many as 453 articles were deposited to arXiv by 24 Southampton researchers while only 21 scholars have deposited 240 articles into the IR. Xia states that this “may indicate that the faculty authors are not as interested in working with their IR as with a SR.” I’d say that’s pretty obvious. The SR is going to reach their target audience far more directly, why duplicate the effort for what will likely be far less payoff? What interests me more is what Xia discusses next - the lack of quality of both the content and metadata in the IR itself. He mentions that a large number of the materials in the IR have likely been deposited by a third party via mediated archiving. He is forced to reject his hypothesis by stating that the most enthusiastic SR depositors tend to generate much less, if any, of their own IR content. Additionally, nearly half of the IR contributors have never worked with arXiv. Thus, no correlation is shown between SR and IR use.

However, what I find to be the most interesting finding in the study is the hilariously awful functionality of Southampton’s IR. Nearly 62% of the articles in Soton are abstract-only, and “some” have links to a full text on an external source. Amazingly, “of the 468 items with an external link to the full text, more than half (253 items) have a dead link.” How astoundingly useless. Xia mentions that this may be caused by indifferent authors who feel forced to self-archive under the pressure of institutional requirements, and who do not follow exact archiving procedure and do not load a full text. Additionally, mediated deposit does not appear to be the final solution, as Xia points out that “it is unrealistic to expect a third-party depositor to read all articles before uploading them and filling out metadata information,” which leads to incorrect, missing, or vague subject terms. He correctly states that mediated depositing can hardly be called “self-archiving.”

He concludes by stating the obvious: when an article has been deposited in one repository, the author(s) will be hesitant to make it available in a second. Also, energetic participation in an SR does not necessarily mean the same in an IR, and vice-versa. He thoughtfully says “[p]ersonal interest in an IR means much more for the development of institutional repositories than being obligated by mandate requirement.” Deposits of abstract-only items with missing metadata and broken links will hardly prove useful in either the short or long term.

This article, while seemingly a simple investigation into the number of deposits of a single group of researchers, really gets to the heart of many of the problems IRs face. Unfortunately, it really makes me more concerned for the future - a mandate alone isn’t the answer, nor is mediated deposit. It’s a larger solution that involves manpower, dedication and personal involvement, something I fear is largely lacking with regard to IRs. Especially with the huge budget issues my state is facing, I can hardly expect any new hiring to take place to increase the number of people involved in scholarly communication. It’s a daunting challenge.

Tags: 1 Comment