Abstract
Statistical analysis and comparison of library print and online holdings are used to reduce shelf area and inform index back-file purchases. Nikki DeMoville of California Polytechnic State University’s Robert E. Kennedy Library combined Counting Online Usage of Networked Electronic Resources (COUNTER) JR2 Access Denied usage reports, holdings information, and print circulation data to identify journals exhibiting a high desirability for online access and low print circulation. In this presentation, DeMoville shared how further combination of this data with ScienceDirect quote lists for back-file journal packages and Web of Science metrics could be used to support the efficient de-selection of print journals, allowing the library to recover the maximum amount of physical space while also acquiring backfile access to those titles most frequently needed in electronic format. The presentation detailed the workflows and comparative methodology used to reach collection management and development decisions for journal coverage, with an emphasis on before and after metrics.
INTRODUCTION
In 2013, California Polytechnic State University hired an architecture firm to develop a 5–10 year master space plan for its Robert E. Kennedy Library in San Luis Obispo; the new plan included recommendations for widespread repurposing of space currently devoted to book and journal shelves. Changes in budgetary and spending practice also came into effect the same year, eliminating the rollover of funds to subsequent years. The result was the one-time availability of $125,000 for a collection weeding and development project.
Under considerable time constraints and using minimal staff and existing technology and tools, the library’s Electronic Resources Coordinator, Nikki DeMoville, and a handful of staff used evidence-based methods to recover 600 linear feet of shelf space and expand online access and improve discovery to content identified and targeted through the project.
The specific goals of the project were:
To clear space in the library stacks by removing appropriate print journal runs in an environmentally friendly manner
To fill gaps in partner organizations’ holdings wherever possible
To spend on content with proven need and to provide better access and discovery to such content
To submit all related invoices within six months of project start
To use existing staff resources
To remain evidence-based throughout all processes
METHODOLOGY
A team of California Polytechnic State University staff, overseen by the Director of Information Resources and Resource Sharing, and comprising the electronic resources coordinator, information resources coordinator, serials coordinator, and a reference specialist, began the project by analyzing COUNTER Journal Report 2 (JR2) as supplied by publishers and aggregators. This report measures instances where access is denied to full-text articles by month, journal, and category. The Team used JR2 to identify journal titles and volumes and issues that library patrons had requested through online interfaces and utilities but not been given access to because the content item had not been licensed by the organization. This denial of access is often also referred to as a turnaway.
Several aggregator and publisher journal and package lists were then analyzed to gauge the correlation of available back-file titles to those titles determined to be in high demand through the JR2 analysis. Ultimately, Elsevier’s ScienceDirect was chosen for the content list as it demonstrated consistently high usage, did not require new licensing for back-file content, correlated strongly with the library’s print holdings, and provided clean, readily available usage and holdings data. In addition, the library had significant front-file holdings through ScienceDirect, which made it possible to easily identify back-file turnaways.
The team then proceeded to collect data from multiple sources in support of further analysis and correlation. Data from Elsevier was used to provide online title lists, back-file pricing, usage, and front-file holdings. Non-Elsevier back-file online holdings were identified through ProQuest Serials Solutions data, and Thomson Reuters Web of Science data was used to help gauge journal impact and discoverability in accordance with extent of indexing. Print holdings and circulation data were taken from the library’s integrated library system, Innovative Interfaces Millennium, and physical holdings overlap with partner organizations was determined from a report compiled from the shared Western Storage Trust distributed print journal repository (WEST).
From the data collected, the team compiled a tool in Excel to aggregate data elements and data sets and formulate a journal-title and package identification method across subject areas of primary interest to the University. The tool calculated total package turnaways, percentage of journal content with turnaways within a package, median number of turnaways both within a package and across ScienceDirect, and the highest number of turnaways from a single title, alongside the estimated cost of licensing the package. At the journal title level, the tool also reported ISSN, full title, title change history, print holdings year range, number of turnaways, front-file subscription coverage, local call number, volume coverage and binding, aggregate print usage, and last date of print circulation, as well as the title’s archive status and physical holdings at WEST member libraries.
The tool provided a scalable method for compiling comparable usage metrics across titles and packages. A total of 24 back-file packages were analyzed and compared. From the analysis and comparison, criteria for the addition of online access and the removal of print runs were determined.
Decision criteria for adding online access included a threshold of total turnaways in conjunction with a high percentage of titles with at least one turnaway within a package, as well as a strong overlap with print holdings exhibiting low circulation and use. The degree of current front-file coverage was also factored into the decision, as were the depth of the target back-file and the overall cost of the package. As an example, a package with over 90% of titles reporting at least one turnaway, a total turnaway count of nearly 2,000, and a highest single-title turnaway rate of over 200 turnaways was considered an obvious candidate for licensing. On the other end of the spectrum, a package with fewer than 50% of titles reporting at least one turnaway, a total turnaway count of just over 100, and a highest single-title turnaway rate of fewer than 35 was not considered sufficiently desirable or needed to warrant licensing.
The withdrawal of print runs was gauged from considering how many times physical volumes had circulated, how recent the last date of circulation was, the general currency of the material, along with the potential online back-file access. As an example, a total print circulation of two with a latest circulation date of greater than 10 years was considered sufficient cause for removal of the physical volumes. Title runs with very high or recent circulation numbers would have been considered for retention on the shelves contingent on performance in the online format and preservation status within WEST, but in actual practice none of the titles under consideration met these criteria, and all were withdrawn as a result. It should be noted that subject areas, such as architecture, with a heavy preference for print formats were excluded from the project in advance.
RESULTS
In total, 24 back-file packages were evaluated, resulting in the selection of 11 packages and 2 individual titles, and 600 linear feet of shelf space were recovered through the weeding of print journal runs based on the analysis performed. This led to expanded online access for researchers and students to journals of current interest, aligned with the university’s main research and teaching areas, as well as potential repurposing of shelf and library space.
To ensure maximum discoverability and use, upon licensing of the chosen titles and packages, each was made active to the full extent possible in the many discovery and access interfaces licensed by the Library. This included activation in the following library-service provider systems and tools:
ProQuest Serials Solutions: KnowledgeWorks knowledgebase, 360 Resource Manager electronic resource management system, Summon discovery service, 360 MARC
ExLibris: SFX link resolver
Innovative Interfaces: MARC record loading to Millennium integrated library system
RapidILL: Electronic holdings refresh with flag set to indicate availability for lending
To align space availability with the 5–10 year master space plan, journal volumes identified for removal were pulled from the library shelves, their holdings records were suppressed or deleted, and all material was offered to partner organizations for consideration in filling collection gaps. Volumes that were not claimed were recycled, and interlibrary loan holdings were refreshed to reflect the Robert E. Kennedy Library’s updated print journal coverage.
In effect, the addition of in-demand online journals and removal of corresponding print volumes was an important step in transitioning the library collection to e-preferred. Overall, the 5,399 turnaways reported for online journal content from 2013 to 2014 were answered by 3,367 successful article requests in 2014 to 2015 from the journal coverage added to the collection through the project. The effect of adding additional indexing was more nuanced; early data suggested that the University saw increased back-file usage across all publishers in social sciences journals, while no noticeable increase in usage was seen in the sciences.
FURTHER OUTCOMES AND APPLICATION
In addition to the tangible results listed above, pitfalls and preferred assessment sources were identified through the project. An important outcome of the project was that it demonstrated a constant need to be careful of duplication in considering journal indexes. For example, Science Citation Index and BIOSIS Previews showed heavy overlap of journal content when compared against the library’s holdings.
Also of note was an often-encountered lack of depth in back-file coverage in indexes, as well as in the Library’s physical holdings records. Experience gained through the project pointed to the following tools as providing the best match rates for the subject areas of primary interest to the University: Science Citation Index and Social Science Citation Index for content reaching back to the mid 1950s and Zoological Record for content from the 19th century.
The usage reports analyzed may likewise pose challenges, as publishers and aggregators transition from COUNTER3 to COUNTER4.Footnote1 For example, COUNTER Journal Report 1a, useful for the fact that it separates out back-file usage, has now been replaced by COUNTER Journal Report 5, which results in collation of usage on journal content older than 15 years.
Usage reports may also work as quick tools for identifying front-file gaps, as well as back-file gaps, and may be used in conjunction with physical trace measures to determine journals where electronic is preferred by the user population. For example, COUNTER Journal Report 2 provided a reliable indicator of interest from the library user spanning the entire life of a journal. However, it was noted that this indicator required low effort from the user and that interlibrary-loan requests might be more indicative of demand. It was further noted that COUNTER Journal Report 2 data was most likely to be useful when drawn from high-traffic sites, such as publishers or platforms where the library already licenses considerable content.
Finally, it was found that the usage data analyzed could be used to identify access issues on publisher and aggregator platforms (e.g., scheduled and unscheduled system down time). This offers insight into overall platform performance and its impact on library users.
CONCLUSION
The statistical analysis and comparison of library print and online holdings performed as part of the California Polytechnic State University Robert E. Kennedy Library’s project to reduce shelf area and inform index back-file purchases proved a reliable, extensible method for current and future needs. Over 600 linear feet of shelf space were recovered for repurposing as space for group study, faculty–student collaboration, campus partnerships, or other needs, and the use of COUNTER usage reports, holdings information, and print circulation data to identify in-demand journals and low print circulation allowed the Library to convert thousands of turnaways to online content not previously available through the library’s electronic collections into thousands of successful article requests. Interlibrary loan processes were streamlined with the conversion to online formats as the need to manually retrieve and scan print journal volumes was reduced. The methodology and tools elaborated through the project made it possible to establish selection and weeding criteria and track the associated metrics in a template fashion that can be reused throughout the Library’s master space plan and into the future.
Additional information
Notes on contributors
Nikki DeMoville
Nikki DeMoville is Electronic Resources Coordinator, California Polytechnic State University, San Luis Obispo, California.
Aaron Wood
Aaron Wood is Senior Manager, Product Development, Ingram Library Services, LaVergne, Tennessee.
Notes
1. Project COUNTER, “The COUNTER Code of Practice for e-Resources: Release 4,” http://www.projectcounter.org/r4/COPR4.pdf (accessed on May 18, 2015).