Spoilt for Choice: A Comparison of Two Aggregations and the use of ScholarlyStats and Ulrichs Serials Analysis System
The proliferation of online databases supplying aggregations of journal articles to support undergraduate courses means that academic library staff need to exercise care and maintain a watching brief on the performance of these databases. In 2006, the University of Melbourne Information Services moved from one large aggregated database to a competitor, after several years of monitoring coverage content and date ranges.
This paper will define the term aggregation as used in the paper and provide details of each aggregator package before going on to discuss the rationale behind the evaluation and the procedures and tools used in the evaluation process (Ulrich’s Serials Analysis System (USAS) and ScholarlyStats (SS) from MPS Technologies).
In 2006, I became the University of Melbourne’s CAUL Datasets Co-ordinator. In this capacity, I established a working party to conduct the comparison of the databases. To ensure a representative sample of staff, I included liaison librarians and collection managers from a wide discipline range. There was a perceived need to provide value for money and ensure transparency of decision making and involvement of stakeholders from other areas.
The first task was to provide spreadsheets to working party participants. These spreadsheets contained lists of full text titles unique to each package as well as lists of all full-text titles with coverage dates included. USAS was then used to provide subject analysis and to eliminate titles already held online through publisher packages or individual subscriptions.
After consultation with stakeholders, members of the working party returned with lists of titles identified as essential. It was then we turned to SS for usage data for these essential titles. It was then possible to recommend purchase of individual title where usage data indicated the necessity, or to point out these essential titles were not used or little used during 2006.
The paper will close with recommendation to continue this type of analysis regularly to ensure our collections reflect the needs of the users while providing financial managers with appropriate data to ensure accountability and value for money.