Friday, May 1, 2015

SSRN & Institutional Repositories

In a previous post, I discussed using SSRN in lieu of an institutional repository to make faculty publications more accessible.

As mentioned in the previous post, it's ideal to use both SSRN and an institutional repository, but not all schools have institutional repositories (like my previous school), so I was promoting the use of SSRN in that case.

My new school does have an institutional repository, which is great. And we also utilize SSRN to promote faculty scholarship. Inevitably, when we discuss this with faculty, we get the question, "will the institutional repository hurt my SSRN rankings or downloads?"

An article that appeared in the AALL Spectrum in 2012 discusses this very issue. "Librarians have every reason to support the creation of an institutional digital repository (IR). An IR preserves the output of the intellectual life of the school, enables anyone with internet access to enjoy the benefits of the new knowledge, and promotes the institution and scholar by bringing to the foreground their intellectual achievements. Plans for a new IR project within the law school, however, can quickly find such worthy motives swept aside as faculty members invariably voice some version of the following comments: “Won’t posting my articles elsewhere steal downloads away from SSRN? That would lower my rankings in SSRN and perhaps reduce my professional stature.” One can regret that law academics today reflexively cower at the thought of appearing to perform poorly on any new ranking system that crosses their path, no matter how dubious. Even so, there can be no denying that SSRN, or the Social Science Research Network, has earned a respectable cachet among the professoriate. This is a tool they believe they understand and with which they’ve grown comfortable. The proper response, then, is not—however tempting it may be—to point out that ranking by downloads is an easily gamed and essentially meaningless metric. Rather, the more successful strategy appeals to the fact that such fears are based upon a flawed appreciation of how readers connect with scholarship of interest."

When faculty ask if the institutional repository will hurt their SSRN rankings, "[t]he question assumes a fundamentally zero-sum view of readers. James Donovan and Carol Watson explain it best in the article by noting that "[i]n this model, a fixed number of readers exists for any given posted article. If the piece is available in only one place, such as SSRN, then all these readers will access the file from SSRN. By concentrating that limited readership in one place, the article and author enjoy their maximum ranking. Should another version of the article become available, as in an IR, that limited audience becomes split, divided between SSRN and the IR. Every download in the repository signifies a lost download by SSRN and vice-versa. Such folk sociology can be remarkably resistant to correction, not least because it could be true. There is nothing obviously false in the view that multiple versions divvy up a limited audience and consequently that the effect of an IR, aside from all the larger virtues it promises, will be to lower the status of any individual author in the SSRN rankings from what it otherwise would have been. This possibility reasonably motivates faculty to jealously shield their SSRN download statistics from potential dilution by a competing website. But just as the argument is not obviously false, neither is it necessarily true. While the total number of readers of any given work is certainly finite, this fact can lead to the mistaken conclusion that it is therefore also bounded. In other words, if the SSRN and IR copies both get 100 downloads, we needn’t leap to the conclusion that without the IR copy the SSRN downloads would have been 200. There is at least as good an argument that the 100 IR downloads represent new readers who would otherwise not have found the piece at all, yielding a net increase in the audience. Choosing between these competing scenarios cannot be based on mere rhetoric but instead must be based on the facts. It can be shown, we believe, that the zero-sum fear is unwarranted. SSRN and IRs more likely draw from different readerships, meaning that downloads recorded for the repository copy represent not diverted SSRN readers but a new audience for the content. SSRN and IRs do not fight for the same eyeballs, but instead target different populations defined by how readers find their way to the desired content."

The authors go on to compare SSRN and IR download performance over a period of time and find that there is an initial burst of downloads when an article is uploaded to SSRN, but the downloads wane over time and are eventually outpaced by the downloads on the IR.

The authors ultimately advise that faculty should use both.

No comments:

Post a Comment