Moving a Unique Collection to Storage

Improving Access Now and Later

Jennifer A. Maddox Abbott

Jennifer A. Maddox Abbott (maddox5@illinois.edu) is a Collection Management Librarian at the University of Illionois at Urbana-Champaign.

Manuscript submitted March 6, 2020; returned to author for revision June 1, 2020; revised manuscript submitted July 31, 2020; accepted for publication September 15, 2020.

The University of Illinois at Urbana-Champaign Library collected college and university publications (the C-Collection) for several decades without allocating the resources to catalog them. A project to make these items discoverable by patrons was initiated, and tens of thousands of items were added to the online catalog. These items were physically stabilized and transferred to the library’s high-density storage facility. A portion of the collection was also digitized, providing electronic access. Although circulation trended downward, there was no clear indication that materials were less accessible in high-density storage, and new items were discovered that had not previously circulated. Digital surrogates of library material clearly allowed the library to reach a much larger audience, and ideal storage conditions to preserve physical materials long-term combined with electronically available copies appear to be an ideal means for providing greater access while preserving content.

The University of Illinois at Urbana-Champaign (UIUC) Library collected college and university publications (the C-Collection)—bulletins, course catalogs, annual reports, schedules, and other ephemera —from domestic and international institutions for several decades but lacked the resources to catalog these materials. A project to make these items discoverable to patrons was initiated, and during a two-and-a-half-year period, tens of thousands of items were added to the library’s catalog. A significant portion of these materials lacked OCLC records and required original cataloging, which suggests that many of the items are uniquely held at UIUC. In addition to making these physical items available for library patrons, they were physically stabilized and transferred to the library’s high-density storage facility. A portion of the collection was also digitized, providing electronic access.

This distinct collection used a locally created classification scheme (beginning with “C”), making it possible to evaluate circulation data for these items through snapshots from the library catalog. The author sought to explore how this project affected access to the C-Collection, considering the impact of a good presence in the online catalog, in particular when that is the only access point for items held in storage, and how that access compares with the ability of patrons to physically browse library collections. Additionally, because a portion of the collection was digitized through the Internet Archive, there was an opportunity to compare physical circulation with digital access. This paper provides an overview of this large-scale collection management project, plus an evaluation of the accessibility of these materials before and after completion of the project.

Background

The C-Collection was originally housed in the library’s main stacks, which lacked sufficient environmental control, with little attention given to materials’ physical condition. Much of the collection was ephemera, and torn corners, rusty staples, and deteriorating paper were very common condition problems. These publications were classified using a locally created scheme, which grouped items by institution. They rarely had spines wide enough to provide identifying information, and although they were shelved in open stacks, browsing the collection was difficult due to the number of thin items packed tightly onto each shelf.

The C-Collection consists of two parts: items from US colleges and universities (domestic Cs) and items produced by colleges and universities outside the US (International Cs). The distinction between domestic and international Cs was determined by following the rules of the locally created classification system, which provided a clear delineation both through the call number scheme and their physical location in the stacks. Because of this clear delineation, the project team was able to easily retrieve and process materials in the order that they were shelved, systematically clearing the area of the stacks that held the domestic Cs a shelf at a time. Although it is possible that there were errors in classification and it is probable that items were misshelved, the materials were more organized than most projects involving cataloging backlogs.

The international portion of the collection was not included in the scope of the project. It is estimated to be approximately one-third the size of the domestic Cs. These remaining items require foreign language expertise in a variety of languages, and it was not feasible to include it in the project scope. Furthermore, this research focused on the portion of the C-Collection housed in the main stacks that was transferred to storage. A small number of additional items have been classified with a C call number but fell outside the scope of this project and the ensuing research. Items housed in other departmental library collections on campus and items in the stacks permanently shelved separately, such as UIUC yearbooks, do not fall under this definition of the C-Collection.

Literature Review

It can be problematic and more expensive to retrieve materials from storage to retrospectively improve cataloging, and by streamlining several goals into a single project, the library made a greater impact and demonstrated good stewardship of resources.1 Careful planning allowed librarians to combine multiple objectives and to accomplish more with less. “Rethinking workflows as projects rather than business as usual can encourage higher productivity, minimize the number of times items need to be touched, and generally create an environment that rewards accomplishment.”2

In addition to increasing discoverability of a previously uncatalogued collection, the project facilitated the digitization of the collection at a later date by creating the necessary metadata, housed materials in preservation quality envelopes as needed, and moved the materials to a climate controlled storage location suitable for long-term preservation. Space in the main stacks was freed up for other collections housed in that location to expand, making that material more accessible.

UIUC conducted a 2002 space study and found that the main stacks were at 99.65 percent capacity, with some ranges as much as 107.6 percent full.3 Atkins and Weible found that transferring materials to storage reduced errors in retrieval and in shelving. Once shelves were less crowded and materials were no longer piled on the floor or on top of other items, retrievers could more successfully find items and shelvers could shelve materials in the correct spot.4

The most cited concerns when a library transfers materials to a storage location include the time required for patrons to receive materials and the belief that materials in storage are less accessible to researchers.5 Although chance discoveries can certainly occur when shelf browsing, as Barclay explains in “The Myth of Browsing,” this method is less successful than patrons believe.6 There are several hindrances to success: the most in-demand items are most likely to be checked out and therefore not discoverable. An item can only be shelved in one physical location, regardless of the number of topics it encompasses or its interdisciplinary nature. Location of items on the shelf can impact success. “Just as products positioned on the middle shelves of grocery stores sell better than those on higher or lower shelves, books that come to rest on the middle shelves of library stacks circulate more than books above or below.”7 Additional barriers to browsing include shelves overcrowded with books that lack spine labels or dust jackets. This collection in particular did not lend itself to browsing because there were thousands of thin paper pamphlets, some in pamphlet binders, packed tightly on the shelves, and most lacked a spine thick enough for any sort of label or indication of content.

Being unable to physically browse library collections means that quality access through the online catalog is even more important for materials transferred to storage. “Material stored without a reliable record is, for most practical purposes, lost.”8 Shlomo pointed out, “the disadvantages of storage can be somewhat ameliorated by full bibliographic description of the titles stored.”9 For this project, extensive work to ensure catalog access was essential. The library “made a commitment to fully catalog all materials going to the [high-density] facility so that library users can search for materials by all available access points.”10

Not everyone is concerned about the loss of physical browsing or the time required to get requested library materials. One study found that most students liked having books retrieved for them and felt they got the books faster than if they had to locate them.11 At UIUC, “items are made accessible to patrons within 24 hours (excluding weekends and holidays) of their request.”12 A survey of university libraries in Australia and New Zealand found that respondents did not indicate user resistance to remote storage, and noted users’ changing attitudes, that “they have come to accept either that access can be provided by means of a digital surrogate or that delayed access to stored content may be necessary if the original item is required.”13

When making decisions related to the physical shelving and storage of library materials, it is essential to remember that “‘low use’ does not mean irrelevant. In fact, low-use materials are often the unique research materials that most distinguish a library’s collection,” making long-term preservation and online access to these unique materials even more important.14 “Off-site storage is a relocation of existing owned materials, not a destruction or replacement of them.”15 Preservation and access may both improve when an item is transferred to storage. When a book is sent to high-density storage, it “does not become forever unavailable or undiscoverable. Thanks to existing and emerging online search tools, books that go off site in the digital age are actually more discoverable than they were sitting on the shelf in the predigital world.”16 Respondents to a survey conducted by Priddle and McCann described how low-use collections transferred to storage began to circulate once they were well-described and discoverable in the online catalog. “The irony is that, once materials are described and accessible, even predicted low-use collections can become desirable for researchers.”17 Burton and Kattau found in their study that transferring materials into non-browsable storage had “not been an impediment to discovery and access,” seeing more than five thousand items circulate for the first time.18

The Project

A report was generated that showed all cataloged items with a C call number that fell within the project’s parameters. Although the C-Collection was physically browsable, most of the collection was only discoverable in the catalog through a single collection-level record. A total of 9,293 items were cataloged, while the collection was estimated to include 45,000 items.

The library’s Collection Management Services (CMS) unit executed the C-Collection project. It began in July 2012 and was completed in April 2015. It took thirty months to complete, with a six-month break in fall 2014 to focus on more time-sensitive projects. A team of four full-time equivalent (FTE) academic hourly project staff retrieved the materials from the main stacks. They stabilized and/or housed almost every item, which consisted of placing them in preservation-quality envelopes in most cases. They added bibliographic records for items lacking an online presence, and when necessary, created bibliographic records for items that were more straightforward or traditional. Two staff members in the unit devoted half of their time to the project, answering cataloging questions and handling the more complicated original cataloging. The items were then transferred to UIUC’s high-density storage facility, which is located on campus, one mile from the main library.

Ultimately, due to the large number of slim pamphlets, the collection was found to contain 63,652 items, which is 41.4 percent greater than originally estimated. At the beginning of the project, more than eighty-five percent of the collection (54,359 items) was essentially inaccessible through the catalog (see table 1). Of the 54,359 items added to the catalog, 26,966 had copy available in OCLC, and 27,393 items required original cataloging (see table 2).

Method

Before the project began in 2012, the team ran a report for all items with the local call number prefix C that fit within the project’s scope to identify how many items in the collection were actually represented in the catalog. The team ran the same report again in 2018 after the project was complete. The library began using Voyager as its Integrated Library System (ILS) in 2002 and migrated to Alma in 2020, so consistent circulation data can be evaluated and compared beginning in 2002.

The project ran from July 2012 to April 2015, with items being continually processed and transferred to storage throughout that period. A small selection of items was digitized at this time, including UIUC publications and pre-1923 items, which could be made available in full view. Although the ideal project would have digitized all items in the collection as they were processed for storage, the resources were not available at the time and most of the items were digitized at a later date through a different project, with most of the digitization completed by 2016. Due to the large-scale nature of this project, tracking exact dates of transfer of individual items was not feasible, but estimates made using the data available were accurate enough to explore circulation trends. For this research, circulation comparisons used the reports run in 2012 and again in 2018, comparing pre-project circulation (2002 through 2012) and post-project circulation (2013 through 2018).

Any items lacking a record in the online catalog had not circulated since 2002 when the current ILS was implemented. The library began barcoding materials in the mid-1990s, and the decision was made to only barcode materials in the main stacks at the point of circulation to hold down costs.19 If a patron wanted to check out an item that was discovered in the main stacks through shelf browsing and it was not in the online catalog, the item was barcoded and a brief record was created in the catalog at that point. For this reason, all uncatalogued items were included in the counts of items that never circulated.

Results

Comparing Circulation

The number of items that circulated one or more times from this collection was very low. Before the project began, 1,044 items (1.6 percent) circulated, and 62,608 items (98.4 percent) were never checked out. After the project was completed, 329 items (0.5 percent) circulated, and 63,323 items (99.5 percent) never circulated (see table 3). Of those 329 items that circulated in 2013 to 2018, 225 (68 percent) had not previously circulated. The remaining 54,134 newly cataloged items had not circulated as of 2018.

Considering only the number of items that circulated and not the total number of circulations that occurred, provides a partial picture of usage. Additionally, the date ranges for pre- and post-project are not equal and comparing the number of items that circulated may not be the most useful view of the data. The total date range is seventeen years, with eleven years of pre-project data and six years of post-project data. The author more closely examined the average circulations within the ranges. From 2002 to 2018, this collection saw a total of 2,183 circulations. From 2002 to 2012, there were 1,735 circulations of the 1,044 items. From 2013 to 2018, there were 448 circulations of the 329 items. The collection had an average of 158 circulations annually before the project, and an average of seventy-five circulations annually after the project (see table 4), leading to a 52.5 percent decrease in circulation.

Circulation of the Digitized Collection

The library’s Digital Content Creation unit digitized 7,836 C-Collection items through the Internet Archive, and 6,553 of those fell within the scope of this project. Of those, fifty-three items (0.8 percent) circulated between 2002 and 2012, and nine items (0.1 percent) circulated between 2013 and 2018 (see table 5). The percentage of the digitized collection that circulated (albeit an extremely small sample size) was slightly lower than for the C-Collection overall.

The digitized collection had 101 total circulations for the entire date range of 2002 to 2018, with eighty-nine circulations taking place prior to digitization and twelve circulations taking place subsequently. On average, the portion of the collection that was digitized had eight circulations per year before digitization and saw two circulations per year after digitization (see table 6).

Electronic Access versus Physical Circulation

Data has also been collected on electronic access to the digitized portion of the collection through the Internet Archive. The Internet Archive calculates a view as “one action (read a book, download a file, watch a movie, etc.), per day, per IP Address. For each item page, using multiple files or accessing from multiple accounts in a single day counts as one view.”20 The item with the fewest views received four, and the item with the most views received 9,111. The digitized collection of 6,553 items had a total of 1,736,128 views, with an average of 265 views per item. The fifty-nine items that circulated were viewed 23,927 times, with an average of 406 views per item, and the 6,494 items that did not circulate were viewed a total of 1,713,201 since being digitized, which is an average of 264 times per item (see table 7).

Figure 1 shows the distribution of views for the digitized C-Collection, comparing items that had circulated with items that never circulated. Of note, more than ten percent of the circulated items also had more than 1,000 views, versus less than three percent of the collection that never circulated being viewed more than 1,000 times. The data is also provided in table 8.

An examination of the fifty-nine items that physically circulated revealed that the nine items that circulated after digitization were viewed for a combined total of 2,007 times, with an average of 223 views per item. The fifty items that did not circulate after digitization were viewed a total of 21,920 times, for an average of 438 views per item (see table 9).

Discussion

When considering the number of items that circulated, plus the total number of circulations, before and after transferring the C-Collection into storage, it is difficult to draw the conclusion that the added discoverability in the library catalog significantly increased access to the physical collection. Circulation numbers were very low both before and after the project. Deutch reported at the time that print circulation was trending down nationwide.21 A study by Reeves and Schmidt saw an increase in the usage of electronic journals and a “dramatic decrease” in usage of print journals, and Cook and Maciel found that patrons prefer electronic resources for convenience and ease of use.22

It is encouraging to see that sixty-eight percent of the items that circulated post-project were from a group of resources that had not previously circulated. If this were a project with a singular goal and no other benefits beyond immediate circulation, it would be hard to justify the resources to catalog 54,359 items for patrons to discover 225 of them. In a situation with less constrained resources, an argument could be made that the discovery of those particular items may have a research-impact potential that extends beyond being quantified by a simple statistic. It is clear that “assessing the return on investment, or cost/value ratio, for cataloging is a difficult proposition at best.”23 Hider pointed out that determining the cost of the work is fairly straightforward, but “determining the corresponding benefits, in monetary terms, is much less straightforward and has rarely been attempted.”24

Resources were allocated and justified because this project had the additional benefits of clearing space in the main stacks, storing the collection in better preservation conditions, physically stabilizing items for long-term preservation, and creating metadata to aid discoverability and to facilitate later digitization. Although the average circulations per year decreased post-project, this outcome does not suggest that the project was unsuccessful. For collections in the library’s storage facility (i.e., prime preservation conditions), the priority is materials with a lasting research value that do not circulate heavily. For this reason, it is not a concern that these materials have not been used during the relatively short time frame.

When comparing the circulation of library materials shelved in open stacks versus storage, a key variable is electronic access. Once an item has a digital surrogate available to patrons, it is to be expected that the physical copies will not circulate as heavily. In this way, more patrons have access while also preserving and reducing damage to the library’s physical copy. Although the sample size was small, the circulation of the digitized portion of the collection did decrease after digitization. Circulation for the entire C-Collection decreased by 52.5 percent (average of 158 circulations per year pre-project down to seventy-five circulations post-project), but the digitized portion decreased by 75.0 percent (average of eight circulations annually to two circulations annually). Although the sample size is not statistically significant, the trend is in line with expectations—physical circulations decrease when an item is also available electronically.

In addition to physical circulations, there is a correlation between circulation and electronic access. It is notable that the items that physically circulated were, on average, viewed more electronically than the items that did not physically circulate. This finding seems to indicate that the items of most interest were of interest in either or both available formats. But the most impressive findings were the number of electronic views for this collection via the Internet Archive. The 6,553 items had more than 1.7 million views, and so expanded the reach and research value of this collection.

Conclusion

This research closely examined the circulation and digital access trends of a discreet collection to better understand the difference in access between library materials in open, browsable stacks that lack bibliographic records in the online catalog versus materials that are discoverable in the catalog but cannot be browsed in open stacks. Circulation trends were down, but there was no clear indication that materials were less accessible while stored in high-density storage. In fact, new physical items were discovered that had not been previously checked out by patrons, presumably due to the newly created bibliographic records. The metadata also facilitated access to digital surrogates which allowed the library to reach an exponentially larger audience than is possible with physical copies. The combination of ideal storage conditions to preserve physical content long-term and electronically available copies for patron access appear to be an ideal means to provide greater access while also preserving content for long-term use. “In the longer term, institutions will need to adjust cataloging workflows and investments in retrospective projects after carefully weighing the costs and benefits of discoverability and access against investment in operations.”25

Future research would benefit from a larger sample size of material and/or a larger timeframe to compare circulation trends of collections in open stacks and high-density storage, plus examining discrete subject areas in more depth to ascertain trends by discipline or subject area. Lastly, it would be useful to explore a larger sample of digitized material and to evaluate access (i.e., “views”) through alternate avenues such as HathiTrust or Google Books.

References

1. Violeta Ilik, “Off-Site Storage from a Cataloging Point of View,” Serials Librarian 63, no. 3–4 (2012): 350–58; Mary S. Laskowski, “When Good Enough Is Not Good Enough: Resolving Cataloging Issues for High Density Storage,” Cataloging & Classification Quarterly 54, no. 3 (2016): 147–58.
2. Jennifer A. Maddox Abbott and Mary S. Laskowski, “So Many Projects, So Few Resources: Using Effective Project Management in Technical Services,” Collection Management 39, no. 2–3 (2014): 161.
3. Jennifer Hain Teper and Stephanie S. Atkins, “Time and Cost Analysis of Preparing and Processing Materials for Off-Site Shelving at the University of Illinois at Urbana-Champaign Library,” Collection Management 28, no. 4 (2004): 43–65.
4. Stephanie S. Atkins and Cherié L. Weible, “Lost Is Found: The Impact of a High-Density Shelving Facility on a Library’s Collection,” Collection Management 31, no. 3 (2006): 15–32.
5. Jim Agee and Sarah Naper, “Off-Site Storage: An Analysis,” Collection Building 26, no. 1 (2007): 20–25; Donald A. Barclay, “The Myth of Browsing,” American Libraries no. 6–7 (2010): 52–54; Claire Q. Bellanti, “Access to Library Materials in Remote Storage,” Collection Management 17, no. 1–2 (1993): 93–103; Mary S. Laskowski, “High Density Storage: From There to Here and Beyond,” Journal of Academic Librarianship 42, no. 2 (2016): 144–50; Amy Lucker, “Deal with the Devil: A Participatory Model for Off-Site Storage Selection,” Art Documentation: Bulletin of the Art Libraries Society of North America 31, no. 2 (2012): 285–92; Bruce E. Massis, “‘Serendipitous’ Browsing versus Library Space,” New Library World 112, no. 3–4 (2011): 178–82; Shirley Rais, Michael A. Arthur, and Michael J. Hanson, “Creating Core Title Lists for Print Subscription Retention and Storage/Weeding,” Serials Librarian 58, no. 1–4 (2010): 244–49; Elka Tenner Shlomo, “Nicholson Baker Wasn’t All Wrong: A Collection Development Policy for Remote Storage Facilities,” Acquisitions Librarian 15, no. 30 (2003): 117.
6. Barclay, “The Myth of Browsing.”
7. Barclay, “The Myth of Browsing,” 54.
8. Ilik, “Off-Site Storage from a Cataloging Point of View,” 351.
9. Shlomo, “Nicholson Baker Wasn’t All Wrong,” 120.
10. Teper and Atkins, “Time and Cost Analysis,” 50.
11. Miriam Deutch, “Paging a Library Collection: The Brooklyn College Library Experience,” Collection Building 20, no. 1 (2001): 25–32.
12. Mary S. Laskowski and Joseph J. Lenkart, “Curating and Accessing Off-Site Special Collections: Area Studies Materials and High-Density Storage,” Collection Management 40, no. 4 (2015): 217.
13. Paul Genoni, “Storage of Legacy Print Collections: The Views of Australasian University Librarians,” Collection Management 37, no. 1 (2012): 41.
14. Scott Seaman, “Collaborative Collection Management in a High-Density Storage Facility,” College & Research Libraries 66, no. 1 (2005): 26.
15. Agee and Naper, “Off-Site Storage,” 21.
16. Barclay, “The Myth of Browsing,” 54.
17. Charlotte Priddle and Laura McCann, “Off-Site Storage and Special Collections: A Study in Use and Impact in ARL Libraries in the United States,” College & Research Libraries 76, no. 5 (2015): 664.
18. Fiona Burton and Maureen Kattau, “Out of Sight but Not Lost to View: Macquarie University Library’s Stored Print Collection,” Australian Academic & Research Libraries 44, no. 2 (2013): 109.
19. Teper and Atkins, “Time and Cost Analysis.”
20. Internet Archive, “Internet Archive Statistics,” September 30, 2019, https://help.archive.org/hc/en-us/articles/360004650632-Internet-Archive-Statistics.
21. Deutch, “Paging a Library Collection.”
22. Robert K. Reeves and Kari Schmidt, “Radical Relocation: Adapting Print Collections to an E-Centric World,” Serials Librarian 61, no. 3–4 (2011): 416; Colleen Cook and Michael Maciel, “A Decade of Assessment at a Research-Extensive University Library Using LibQUAL+®,” Research Library Issues no. 271 (2010): 4–12.
23. Laskowski, “When Good Enough Is Not Good Enough,” 148.
24. Philip Hider, “How Much Are Technical Services Worth?: Using the Contingent Valuation Method to Estimate the Added Value of Collection Management and Access,” Library Resources & Technical Services 52, no. 4 (2008): 254.
25. Laskowski, “When Good Enough Is Not Good Enough,” 156.