Elisabeth Kaplan
Associate University Librarian
George Washington University
Philip Herold
Research & Learning Director for Agricultural, Biological and Environmental Sciences
University of Minnesota, Twin Cities
Kenning Arlitsch
Dean
Montana State University
On Assessing Institutional Repository (IR) Efficacy Over Time: A Case Study in Context (Kaplan, Herold)
Recent calls for a critical re-examination of IR services have sparked conversations focused on the efficacy of IR as support for open access to scholarly publications. In fact IRs are far more diverse than the conversation has thus far reflected, and measures of their effectiveness need to take this into consideration. In particular, IRs that are well-integrated with institutional goals tend to be more successful. In this session, two former IR managers look back on the initial rationale for establishing an IR at the University of Minnesota, and assess how well that has played out over the past 10 years. We suggest that IRs have and will serve increasingly important functions, particularly for public universities with a mandate for transparency and accountability, and that delivering value to the institution is paramount to any definition of success.
Undercounting File Downloads from IRs (Arlitsch)
A primary impact metric for IRs is the number of file downloads, which are commonly measured through third-party web analytics software. Google Analytics, a free service used by most academic libraries, relies on HTML page tagging to log visitor activity on Google’s servers. However, web aggregators such as Google Scholar link directly to high value content (usually PDF files), bypassing the HTML page and failing to register these direct access events. Conversely, overwhelming and ever-evolving robot traffic renders log file analytics software very difficult to use and usually results in over counting.
This presentation presents evidence of a study of four institutions demonstrating that nearly 60% of IR human-generated activity is not counted by page tagging web analytics software, and proposes a practical solution for significantly improving the reporting relevancy and accuracy of IR performance metrics using Google Analytics.
http://conservancy.umn.edu
http://scholarworks.montana.edu/xmlui/handle/1/9943
Presentation (Arlitsch)