C&RL News October 2018 506 Collection weeding is an important func-tion of any library operation. At its sim- plest, the process of deciding what to remove, and implementing it, is controlled by a library or system serving a single mission and set of objectives. This becomes more complicated when the collection is shared by a group of libraries, such as those within a consortium, that have separate governing structures. The difficulty of such an undertaking is further compounded for collections of electronic resources with varying licensing terms and technical configurations. This paper describes just such a weeding project of an e-book col- lection shared by members of a large library consortium. California’s Community College Library Consortium (CCLC)1 was facing a challenge in 2013 with its shared collection of more than 27,000 e-books. The collection was ag- ing, and some members asked for content to be removed. Reaching a consensus on which titles to weed was a daunting task. In response, CCLC undertook a multiyear project whereby the problem was defined, a plan for addressing it was developed, and steps were taken to resolve it. This resulted in weeding nearly 12% of the collection and laid the groundwork for future collection reviews by the consortium. Background Between 2001 and 2010, CCLC member li- braries shared access to customized e-book collections offered annually at discounted rates by NetLibrary (later EBSCO). Collec- tion packages ranged in size from 1,073 to 3,174 titles. The annual offerings consisted of new and older imprints from approxi- mately 155 publishers. By the last collec- tion, there were 27,622 titles with copyrights spanning 1913 to 2011 (77% were published between 2001 and 2011). As with all CCLC offerings, each member library decided individually whether to buy an annual collection. Seven libraries opted to buy the first one, and 105 colleges bought at least one collection during the ten years they were offered. For many participating libraries, this was their first foray into the world of e-books. By the last collection in 2010, the land- scape for e-book acquisitions had shifted, and many libraries had grown comfortable purchasing and providing access to e-books directly for their own students. In 2011, EBSCO notified CCLC that they would not offer any more shared collections. Tamara S. Weintraub, Brian Greene, and Glorian Sipman Weeding a shared e-book collection Collaboration across a consortium Tamara S. Weintraub is serials and e-resources librarian at Palomar College, email: tweintraub@palomar. edu, Brian Greene is librarian at Columbia College, email: greeneb@yosemite.edu, and Glorian Sipman is technical services/collection development librarian at MiraCosta College, email: gsipman@miracosta.edu © 2018 Tamara S. Weintraub, Brian Greene, and Glorian Sipman mailto:tweintraub%40palomar.edu?subject= mailto:tweintraub%40palomar.edu?subject= mailto:greeneb%40yosemite.edu?subject= mailto:gsipman%40miracosta.edu?subject= October 2018 507 C&RL News CCLC’s approach to shared collection management Colleges that bought into any of the ten col- lections were guaranteed access in perpetu- ity to the licensed content. However, as time went on, not everyone wanted access to some of the earlier content. Newer editions of some titles were added to subsequent collections, some titles became outdated altogether, and others no longer matched individual colleges’ curriculum. By 2013, most imprints were more than five years old. This rapidly aging collection was be- coming a problem, particularly for colleges offering programs and disciplines where the use of outdated or superseded information could pose a health and safety risk, com- promise certification or employability (such as nursing or EMT), or simply jeopardize students’ academic success. Member librar- ies began reporting their concerns to the Electronic Access and Resources Committee (CCL-EAR) of the Council of Chief Librarians (CCL).2 However, due to the access configuration of shared titles in EBSCO’s e-book platform, participants cannot suppress, remove, or “turn off” e-books they share with others, even via their own vendor portals. With the CCLC shared collection, it’s an all or noth- ing situation, where title removals affect all license holders. One option for libraries no longer wishing to provide access to specific titles is to remove catalog records from their ILS, reducing the chance that a given title would be found. Still, the title remains “dis- coverable” by patrons in other ways, includ- ing through EBSCO eBook Collection portals. When the first shared collection was of- fered, the long-term consequences of this configuration, as well as e-book search and retrieval behaviors of patrons, were not fully understood by participating libraries. The CCL-EAR e-book selection policy3 was written to include criteria for adding titles to an e-book collection, but omitted guidelines for deselection except to advise libraries to remove local catalog links to titles they didn’t want. However, as students bypassed catalogs to access e-books directly at vendor sites, this proved to be an insufficient solu- tion to the problem. By September 2013, CCL-EAR agreed it was time to undertake a project to deselect and remove e-books from the shared collection. Project planning Undertaking this weeding project was no small feat, as it involved a few volunteer librarians making collection development decisions that would impact institutions in the largest system of higher education in the United States. To ensure the project’s suc- cess, CCL-EAR approached it in three steps. The first involved exploring the problem, which included identifying the benefits of undertaking the project, proposing ways to address it, and getting approval for the plan from the consortium’s leadership body. The second step involved conducting a comprehensive collection review to identify titles to remove from the collection. In the final step, titles were removed. The first two steps were undertaken by teams of librarians who skillfully represented the interests of other consortium members, and the third was a coordinated effort between the consortium’s director and the collection’s vendor, EBSCO. The first step took place from October 2013 through May 2014, and began with a CCL-EAR Committee task group exploring and defining the problem, and developing a plan for identifying titles to deselect. Although making selection policy recom- mendations was not one of their charges, the task group reviewed the e-book selection policy to affirm that the current criteria was still valid. They then spot-checked titles in all ten e-book collections, examining disciplines, subjects, and publication dates to identify possible criteria to be used for deselection. With this information, the group then wrote a report that served as a combined proposal for a deselection policy and project plan. After review by the entire CCL-EAR Committee, the proposal was adopted by CCL in May 2014.4 C&RL News October 2018 508 Reviewing the collection After the project proposal was approved, CCL-EAR appointed a project leader who began building a review team, starting with the CCL-EAR committee chair issuing an in- vite to consortium librarians. The only crite- ria for serving was that volunteers had to be employed by a CCLC library with access to at least one of the shared collections. It was as- sumed that as current California community college librarians, everyone would possess the credentials necessary to evaluate the col- lection within the parameters outlined in the project proposal. The final deselection review team included 11 librarians from colleges throughout the state and reflected the geographic and curriculum diversity of the CCLC member libraries. The team included full-time and adjunct librarians from all library operational areas, including systems, reference, instruction, serials, acqui- sitions, cataloging, electronic resources, and administration. The review workload was di- vided more or less evenly, with each reviewer assigned roughly 2,500 e-books. Depending on the size of an annual collection, and who had access, librarians were assigned either an entire annual collection to work on or parts of one or more different collections. The project leader next created Excel work- sheets for each reviewer, which contained a row for each assigned title and columns for corresponding metadata to identify, locate, collocate, sort, and record review data.5 The team also drafted review guidelines6 and a coding system that standardized recommen- dations so they could be easily compiled and sorted.7 The guidelines provided instructions on how to apply the criteria for making a recommendation, and for entering the data into the worksheet. The review period was divided into four stages: 1) an initial pass of assigned lists by each reviewer, 2) a follow-up examination, by all reviewers, of any title that the first re- viewer was unsure of, 3) a final look by the broader CCL-EAR Committee of titles recom- mended for deselection by the review team, and 4) an examination of the deselection list by librarians at any CCLC member library, with a chance for them to appeal the recom- mendation. The first stage took place between Sep- tember and December 2014. Reviewers could recommend “Keep,” “Deselect,” or “Unsure.” The last option was also used for identifying titles that received conflicting recommenda- tions from different reviewers (since team members could make recommendations on titles not assigned to them). Reviewers were encouraged to seek input from each other, as well as colleagues at their own or other institutions. This occurred most often for books on subjects or disciplines unfamiliar to a reviewer, or if there was un- certainty about a book’s value to others. In some cases, the entire review team discussed whether to recommend the removal of out- dated titles that didn’t clearly fall into the nar- rowly defined deselection criteria. Examples included superseded travel guides, computer manuals, books on resume writing, and some titles covering medical, psychological, or consumer health topics that were outdated but would not jeopardize somebody’s health, career, or educational success and could be useful for historical research. In these cases, the team referred back to the deselection criteria and reaffirmed that if there were doubt about the utility of an outdated, but harmless, item, the recommendation should be to “keep.” (In its final report to CCL-EAR, the team recommended that another weed- ing effort be made in two or three years to focus solely on these problematic categories.) All worksheets were submitted in early December as planned, and the team leader compiled the results into a single spread- sheet with recommended action. From this initial examination of the entire collection, 3,064 titles, or 11%, were recommended for deselection. Another 512 e-books (2%) were marked “Unsure” and designated for closer examination by the entire team in the next stage. For the remaining 24,046 titles (87%), no further action was taken. The leader sent the “Unsure” titles to the entire team to examine between January 25 October 2018 509 C&RL News and February 6, 2015. In this round, everyone reviewed each title and made a recommenda- tion to keep or deselect. If a reviewer was unsure, they were instructed not to make a recommendation. Recommendations were tallied to determine what action would be taken. For example, a title with six out of eleven recommendations to “deselect” was added to the deselection list. A title with more “keeps” than “deselects” was not. From this second round of reviews, an additional 357 were added to the deselection list. Appeals process For the next phase, the review team’s recom- mendations were examined by CCL-EAR at their February 2015 teleconference. Although a few titles were investigated further, no changes were made. The CCL-EAR Commit- tee chair next disseminated the recommend- ed deselection list to CCLC member libraries via the consortium listserv. They were en- couraged to review the titles and submit “ap- peals,” to restore any they wanted to keep. The communication included an explanation of the action that would be taken regarding appealed titles, and outlined the timeframe for final approval by CCL and removal from members’ EBSCO’s e-book portals. In the end, CCLC member libraries ap- pealed the deselection of 255 titles. Specific reasons either weren’t stated or members indicated the titles were still relevant for their students. Both the original review team and the CCL-EAR Committee re-examined those titles, and after considering factors such as programs offered at those colleges, decided to keep 100 of them. This brought the final deselection number to 3,274, or nearly 12% of the entire collection. Final steps The final list of deselected titles was posted on the CCL-EAR website in May 2015 for CCLC members to access. In preparation for title removals, libraries were advised to re- view and download the file, and take action on removing MARC records and other meta- data from their local library systems. During this time, members freely shared program- ming scripts, files, and suggestions for batch removal of the titles. CCLC’s director worked with EBSCO throughout summer and fall 2015 to coordi- nate removal of the titles. Although originally planned for June 2015, this turned out to be more problematic than originally anticipated be- cause of members’ varying collection holdings and EBSCO e-book site configurations. Finally, on May 27, 2016, CCLC members received noti- fication through the consortium listserv that the titles had been removed. Conclusion Properly maintaining a collection of e-books shared by dozens of libraries is challenging. For CCLC, technological limitations meant that the rights of member libraries who purchased titles in perpetuity needed to be balanced against the disadvantages and potential harm of leaving outdated materials within the collec- tion. In the end, the various groups involved in this project tackled the problem head on by adhering to key principles: developing a plan that was adopted by the membership’s repre- sentative governing body, following the ad- opted plan by conducting a thorough review of the collection and allowing multiple oppor- tunities for feedback from the membership, and broadly communicating the need for and status of the project throughout its duration. CCLC found there was broad consensus to remove 12% of the collection. Another portion of the collection covering specific categories of books was identified as requiring additional scrutiny in a future deselection project. The re- sult of this effort is that California’s community college students have access to an archival e- book collection that is more usable and reliable, and the CCLC has a blueprint for conducting deselection projects in the future. Acknowledgements The authors wish to recognize the contribu- tions of the following people to the deselec- tion project and this article: Sandra Bierdzin- ski, Linda Braiman, Morgan Brynnan, Norman Buchwald, Robert Frouin, Steve Hunt, Carol C&RL News October 2018 510 Hutte, Daniel Kiely, Rochelle Perez, Jean Ping, Sarah Raley, Jean Smith, Darryl Swarm, and James Wiser. Notes 1. CCLC is a partnership of the Council of Chief Librarians (CCL), a not-for-profit organization that represents and promotes the interests of libraries in California’s public community colleges, and the Community College League of California, a nonprofit public benefit corporation whose volunteer membership consists of 72 community col- lege districts in the state. 2. CCL-EAR Committee, appointed by CCL, includes librarian representatives from ten regional areas of California and acts in an advisory capacity to the CCLC, evaluat- ing offers, including ebook collections, and preparing product reviews. 3. Council of Chief Librarians, “Collec- tion Development Policy for the CCL Shared E-Book Collection,” 2009, https://cclibrar- ians.org/sites/default/files/providers/ebooks /EbookCollectionDevelopmentPolicyFinal.pdf. 4. Council of Chief Librarians, “CCL NetLi- brary/EBSCO EBooks Shared E-Book Collec- tions De-Selection Policy and Procedure, May 7, 2014, May 21, 2014 rev,” Ebook Deselec- tion Procedure & Documents, 2018, https:// cclibrarians.org/sites/default/files/providers /ebooks/deselection_procedures.pdf. 5. Council of Chief Librarians, “Sample Reviewer Spreadsheet,” Ebook Deselection Procedure & Documents, 2018, https:// cclibrarians.org/sites/default/files/providers /ebooks/deselection_procedures.pdf. 6. Council of Chief Librarians, “CCL Shared eBook Collections Deselection Task Force Review Procedures, 9/11/2014,” Eb- ook Deselection Procedure & Documents, 2018, https://cclibrarians.org/sites/default/files /providers/ebooks/deselection_procedures.pdf. 7. Council of Chief Librarians, “CCL Shared eBook Collections Deselection Task Force Codes for Review Master Spreadsheet, 11/18/2014 rev.,” Ebook Deselection Proce- dure & Documents, 2018, https://cclibrar- ians.org/sites/default/files/providers/ebooks /deselection_procedures.pdf. Sales through dealers and carriers, street vendors, counter sales, and other non-USPS paid distribution: not applicable. Other classes mailed through the USPS: not applicable. Total paid and/or requested circulation: Average, 10,424 Actual, 10,309. Free distribution by mail: Average, 119; Actual, 94. Free distribution outside the mail: Average, 0; Actual, 0. Total free distribution: Average, 119 Actual, 94 Total distribution: Average, 10,543; Actual, 10,403. Copies not distributed: Office use, leftover, spoiled: Average, 556; Actual, 632. Total (sum of previous entries): Average, 11,098; Actual, 11,035. Percent paid and/or requested circulation: Average, 98.8%; Actual, 99.1%. Statement of ownership, management, and cir- culation (PS Form 3526, July 2014) for 2018 filed with the United States Postal Service, Postmaster in Chicago, Illinois, September 17, 2018. Statement of ownership and management College & Research Libraries News is published 11 times a year (monthly, combining July/August) by the American Library Association, 50 E. Huron St., Chicago, IL 60611. American Library Association, owner; David Free, editor. Second-class postage paid at Chicago, Illinois. Printed in the U.S.A. As a nonprofit organization authorized to mail at special rates (DMM Section 423.12), the purposes, function, and nonprofit status of this organization and the exempt status for federal income tax purposes, have not changed during the preceding 12 months. Extent and nature of circulation. (“Average” figures denote the number of copies printed each issue during the preceding twelve months; “Actual” figures denote the number of copies of single issues published nearest to filing date.) Total number of copies (net press run): Average, 11,098; Actual, 11,035. Total paid/ requested subscriptions: Average, 10,424; Actual, 10,309. https://cclibrarians.org/sites/default/files/providers/ebooks/EbookCollectionDevelopmentPolicyFinal.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/EbookCollectionDevelopmentPolicyFinal.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/EbookCollectionDevelopmentPolicyFinal.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/deselection_procedures.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/deselection_procedures.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/deselection_procedures.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/deselection_procedures.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/deselection_procedures.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/deselection_procedures.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/deselection_procedures.pdf https://cclibrarians.org/sites/default/files/providers/ebooks/deselection_procedures.pdf