date: 2013-11-30 09:23
Will have to figure out a workflow from RStudio to Octopress. Until then, some comments related to my dissertation and a link to the code on RPubs: http://rpubs.com/seancsb/libsoa
For my dissertation 1), I used data from CiteULike to estimate the probability of being able to use Google Scholar to retrieve full text (or open access or OA) scholarly journal articles without the benefit of or access to a library's scholarly databases and journal subscriptions. The purpose was to understand how libraries are fairing in a world where journal articles are increasingly freely available (e.g., as OA) and, jointly, where people, including scholars / researchers, are increasingly using search engines such as Google Scholar. Fundamentally, is it possible for a serious researcher to function without the benefit of access to a library's services?
It is an important question because for the last century, the academic library was the main source of access to scholarly literature (not the only source, certainly). In the last 20 years, though, the web has raised our expectations concerning what should be freely available. And whether right or wrong, the expectation that everything (all data) should be freely available with a mouse click has been increasing. With these rising expectations and with corresponding changes in scholarly information seeking activities, the future of the academic library is subject to quite a bit of change.
Without taking into consideration the context of the study, the results from Part 1 (see link above) seem to suggest that open access journal articles receive more citations than articles that are not freely available. But since the data is based on references saved in CiteULike and bibliometric data collected from Google Scholar, what it more accurately suggests is that, at the very least, CiteULike users are more likely to collect and use articles that are freely available.
The take away, at this point, has interesting implications for academic libraries. The implication is that open access (or simply freely accessible) articles have such a high payoff, given the previously mentioned expectations and changing information seeking patterns, over articles that are not freely available, they will be more likely retrieved and used over articles that are not freely available. The implication for the academic library can be stated as a rhetorical question: why use the academic library if I can use Google Scholar to retrieve open access or freely accessible articles?
There may be no reason to use the academic library if such is the case and if the case eventually offers even higher payoffs (more OA, e.g.). But then, what is the role of the academic library? Well, one important part of the answer comes from where those OA articles originate. That is, where is Google Scholar retrieving these articles from? In Part 2 (see link above), we see that the university (a label I use to describe a variety of hosting services, but most importantly and most predominately, university-hosted subject and institutional repositories) is the main source of open access articles (Green OA). In the 2012 part of this study (which is longitudinal: 2010 - 2012), 145 universities provided 199 open access articles. This accounted for 38% points of the articles that were freely available in 2012 (58% of the articles were OA by 2012).