Evidence-based library book shifting

Evidence-based library book shifting Joshua Lambert aUniversity Libraries, Missouri State University, Springfield, Missouri, U.S.A. Joshua Lambert Duane G. Meyer Library Springfield, MO 65897 [email protected] https://orcid.org/0000-0003-1920-7549 Joshua Lambert is a Librarian and Associate Professor at Missouri State University. He manages government documents, provides reference assistance, teaches courses for credit, provides library instruction, and gathers and analyzes data for the library and other researchers. Evidence-based library book shifting The Missouri State University Libraries Access Services implemented an evidence-based process to move many books and shelves. That process included measuring the collection, performing quality control checks, creating a plan, and defining regular waypoints to keep the shift from deviating from the plan. This article describes the move, the ideas behind it, and the spreadsheet that provided evidence to make informed decisions. Keywords: stack management; library relocation; book shift; library shelving; Introduction How can library staff accurately shift many books and shelves? What quality assurance activities should library staff implement to make sure they do not need to shift books twice? A renovation required the Missouri State University’s Duane G. Meyer Library to shift many books. Due to logistics related to moving the shelving, library staff needed to shift books forward at times and backward at other times. They also needed to shift multiple parallel ranges simultaneously. To assure a quality shift, library staff wanted to know, before they started moving books, the future shelf location of those books. Missouri State University (MSU) is a doctoral/professional university under the Carnegie Classification system in the United States. It has over 22,000 students, over 80% of whom are undergraduates. The Dwayne G. Meyer Library contained over 800,000 books and library staff needed to move the books between the LC call number range of Q-Z. Library staff measured 97,833 inches (8,152.75 feet) and had to place them on 135,240 inches of shelf space. The shift took many hours of physical labor and had a strict deadline. In addition to the complicated shifting mentioned above, librarians decided it would be better to split the set of books into eleven collections so they could vary the amount of empty space on shelves. This meant that librarians could not use a simple average to know how many books they should shift to each shelf. Even if they used a simple average, it would not guarantee that small mistakes would not lead to large discrepancies by the end of the shift. Library staff determined that measuring books would provide the most accurate results. The shift did not lessen the normal staff workload; therefore, efficiency was important during the process. Staff recorded measurements on paper and then entered them into a spreadsheet. This article uses the following terminology to refer to library shelving. A shelf is a horizontal piece on which books sit. A section is a vertical group of shelves of equivalent size stacked one above the next. A range is a group of sections attached to each other forming a line. Libraries typically place two ranges back-to-back and have aisles between the ranges. Literature Review Melvil Dewey advocated for a separation of books based on size, to fit more books in various spaces (Dewey, 1887). Early published literature related to libraries moving books in the United States comes from the 1930s. The Yale University Library Gazette contained an article in 1930 that explained counting shelves, estimating growth, numbering ranges, measuring shelf sizes, and labeling boxes of books to move to specific shelves (W., 1930). Two years later, Sheetz described a similar process used by the Pennsylvania State Library (Sheetz, 1932). The first mention of finding the average inches of books per shelf came from Ansell, whose measurements and calculations ran “into several hundreds of foolscap folio pages” (Ansell, 1935). Fremont Rider wrote about compact book storage and volume filled by various methods of shelving (Rider, 1949). Hammer mentioned the expansion available per shelf ratio which later authors more commonly called “fill ratio” (Hammer, 1960). An early book about moving libraries is, Moving a Library (Kurth & Grim, 1966). It contained the mathematical ideas from previous library literature as well as additional formulas specific to their move. It included a discussion of measuring linear feet of books, taking samples to estimate linear feet based volume count, sampling error, occupancy ratio (later called fill ratio), and even using the “compound interest formula” to find the annual percent of change of volumes. They modeled future need for shelf space based on past data. In the 1950s and 1960s, librarians collaborated with academics from the field of operations research to better optimize many library operations. According to the School of Operations Research and Information Engineering at Cornell University, “Operations Research (OR) applies the scientific method to the management of organized systems in business, industry, government and other enterprises” (2018). The entire issue of Library Quarterly volume 42, issue 1 in 1972, was dedicated to operations research in libraries. Leimkuhler, Cox, Gupta, and Ravindran wrote articles related to the storage of books on library shelves and were all from the Purdue School of Industrial Engineering and worked with Purdue University library personnel. According to Leimkuhler, “An early study sought to develop a precise measure of the space requirements for various shelving arrangements and to devise an algorithm for finding that arrangement which would minimize the wasted space” (Leimkuhler, 1968). Cox wrote his dissertation related to space optimization in 1964 and that same year, Leimkuhler and Cox condensed that work into an article (Cox, 1964; Leimkuhler & Cox, 1964). They determined the efficiency of shelving little-used material according to height intervals and gave suggestions as to how many height intervals libraries could use for maximum efficiency. Ravindran and Gupta revisited these ideas in 1970 and 1974 using a shortest path network model (Gupta & Ravindran, 1974; Ravindran, 1970). These operations research studies sought to solve a different problem than this article does, but there are a few similarities in the mathematical methods. In 1975, Bommer wrote an article criticizing operations research studies of libraries (Bommer, 1975). He said they were often too costly, too mathematical for most librarians, and impractical. The influence of operations research on library literature since the early 1980s has been minimal. In 1983, Kurkul published an article with a page containing seven mathematical formulas (Kurkul, 1983). Those formulas were not new at the time, but Kurkul used all of them and presented them in a more precise and understandable manner than previous publications. After this article, most of the publications have contributed by explaining new tools or processes as computers and spreadsheets have become integral parts of library operations. Moving Library Collections: A Management Handbook and Library Relocations and Collection Shifts are two more books related to moving libraries that include similar math formulas to Kurkul above (Habich, 1998; Tucker, 1999). Habich, on page 15, discusses segmenting “the collection into internally homogenous groupings.” The process described in this article uses a similar idea below but calls them collections rather than segments. Steven Carl Fortriede’s book and accompanying spreadsheets provide both information and the tools needed to make a library move successful (Fortriede, 2010). His work uses the term “waypoint” and describes it as “a defined spot in a collection designated by the call number of the book to be filed immediately after the waypoint.” In 2011, Castro explained different tools and methods used for a large shift at the University of Miami (Castro, 2011). After Fortriede’s book, Guise, Lindsay and others used his books and spreadsheets to accomplish library moves (Guise, 2016; Lindsay, 2017). Chappell and Krack created websites explaining how their library shifts books (Chappell, 2006; Krack, 2014). Taylor helpfully created a select bibliography related to moving libraries (Taylor, 2015). There are two computer program projects that relate to book shelving. The Collection Shift Estimation and Visualization Tool created by the Harvard Library Lab has the potential to decrease shift planning significantly, but it has requirements that many libraries do not meet (Harvard Library Lab, 2013). It requires that MARC records for the books list the correct number of pages. It also requires that the books use the Library of Congress classification system. The current article adds several ideas and methods to the literature mentioned above. First, it distinguishes between imposed and automatically calculated fill ratios. Second, it explains and demonstrates how to make use of those fill ratios in a spreadsheet (Lambert, 2022a). Third, after measurement data is in the spreadsheet, it provides waypoints for every shelf (Lambert). Lambert created and described a Python program called bookshift.py, which introduced the ideas of imposed and automatically calculated fill ratios (Lambert, 2022b). While such a computer script may work well for some people, it requires a Python programming language environment to execute, and that may prevent some library staff from using it. The spreadsheet highlighted in this article is a more practical solution for many librarians. Research Question The author's primary question was, "On what shelf should each book reside after the shift is complete?" He wanted this information before anyone shifted a single book. Library staff answered the question above by determining how many inches of books they should place on each future shelf. They further wanted to regularly check their progress by determining which book they should place first on every range. This second step established waypoints often enough that library staff could correct inconsistencies with minimal extra effort. While the library literature above provided ideas to build on, it did not provide methods granular and dynamic enough to meet the library’s needs. This article describes an evidence-based method for shifting and provides best practices for other librarians to follow. The “evidence-based-shifting.xlsx” spreadsheet is similar to, but not the same as, the method originally used by MSU library staff. The shelf measurements in the spreadsheet are approximately the same as the original measurements. The spreadsheet is linked to this paper as a dataset. Methods The outline below describes the steps of the shifting process. Some words and phrases will be defined later. 1. Determine what books to move. 2. Determine where to move the books. 3. Divide the books into collections. 4. Record characteristics of books as they are on the current shelves. 5. Record characteristics of shelves in the future configuration. 6. Record characteristics of collections. 7. Adjust table sizes and cell ranges as appropriate in the linked spreadsheet. 8. Perform quality control checks. One of the first steps in a book shift is for library staff to identify which books to shift. MSU’s Head of Access Services received the news that he must coordinate the move of books in the Library of Congress call number range Q-Z. At that time, he also negotiated where to place future shelving. The next step in the shifting process is for a library to decide if they want to divide the books they will shift into multiple “collections.” For this article, a collection is a series of books that cover a contiguous call number range and that have the same fill ratio. This definition breaks down when books from two collections are on the same shelf. The spreadsheet accounts for this complexity, but this article does not explain the details. If a shelf has books from two collections on it, then the fill ratio for that shelf will be somewhere between the fill ratio for each collection. The “book_measure” column of the “data_after” worksheet displays this for the last shelf of a collection. If library staff believe that the number of books in some call number ranges will increase faster than others over time, they can allocate more growth space on those shelves. The methods explained in this article assume a library wants to divide the books into multiple collections. MSU chose to separate the Q-Z books into twelve collections. The “collections” worksheet displays information about those collections. The process described in this article is distinct from past shifting methods, in part, because it uses the idea of imposed fill ratios. An “imposed fill ratio” is not a calculated fill ratio but is something library staff approximate in their mind based on best guesses, intuition, or other information. Library staff may do this because they know that one collection will grow faster than another. Library staff at MSU wanted to impose fill ratios for some collections, but not all of them. The spreadsheet calculates the rest of the collections’ fill ratios, which are the non-imposed fill ratios. For example, if library staff divide the shift into four collections and want to impose fill ratios of 0.65 and 0.7 on two of the collections, the other two may have fill ratios calculated to be 0.74. For further mathematical details, see Lambert’s presentation (Lambert, 2016). To provide a predictable and flexible book shift, library staff recorded up to four characteristics of each shelf: the horizontal measure of books, section increment, call number, and collection number. Library staff must measure the books on each shelf. The “string method,” referred to in many resources about library shifting, is not compatible with the method explained in this article. The section increment column in the spreadsheet indicates the first shelf of a new section. The call number indicates the last book on the first shelf of the first section of a range library staff may later use it as a waypoint. 1 MSU Library staff used paper forms such as the example in Figure 1. That form’s layout visually displayed sections, so they do not have to be recorded separately. Library staff only need to record the call numbers at the beginning of a range. Library staff may choose to add collection numbers as they measure shelves, but MSU staff wrote them after they completed recording measurements. With two people working together, measuring and recording progressed quickly. One person used a yardstick to measure and call out measurements for each shelf. The other person recorded the measurements. Since library staff chose to use paper, the numbers were later input into the spreadsheet, along with the section increments, call numbers, and collection numbers. When performing data entry, MSU library staff found it was quicker to first type a range of measurements in one column. Second, input section increments. Third, input call numbers, and lastly, input collection numbers. Collection numbers span many shelves, and the number can be copied down hundreds or thousands of rows at a time. MSU Library staff created and used quality control checks. The worksheet named “BeforeShelfByShelf” will mimic the layout of shelves and is a quick visual way to do quality control of measurements and data entry. Furthermore, library staff may 1 MSU library staff originally wrote call numbers for the first book on the first shelf rather than the last book on the first shelf. Later spreadsheet formula changes required the last book rather than the first book on the shelf to be recorded. The call numbers were not updated. Users of the spreadsheet can ignore this historical fact. Cox, J. G. (1964). Optimum storage of library materials [Ph.D. Dissertation]. Purdue University. Dewey, M. (1887). Library notes: Improved methods and labor-savers for librarians, readers and writers. Library Notes, 2(6), 95–97. HathiTrust. Fortriede, S. C. (2010). Moving Your Library: Getting the Collection from Here to There. American Library Association. http://www.alaeditions.org/web-extramoving-your-library Guise, J. (2016). Eckhardt-Gramatté Music Library: Chronology of a Building Project. CAML Review / Revue de l’ACBM, 44(2). http://caml.journals.yorku.ca/index.php/caml/article/view/40267 Gupta, S. M., & Ravindran, A. (1974). Optimal Storage of Books by Size: An Operations Research Approach. Journal of the American Society for Information Science, 25(6), 354–357. Habich, E. C. (1998). Moving Library Collections: A Management Handbook. Greenwood Press. Hammer, D. P. (1960). Operation book shift. College & Research Libraries, 21(5), 393–394. Harvard Library Lab. (2013, September). Collection shift estimation & visualization tool. https://osc.hul.harvard.edu/liblab/proj/collection-shift Krack, R. (2014, August 12). Collection management. Rutgers University Libraries’ Staff Resources Access Services. https://web.archive.org/web/20141216213546/http://www.libraries.rutgers.edu/r ul/staff/access_serv/coll_mgt/coll_mgt.shtml Kurkul, D. L. (1983). The planning, implementation, and movement of an academic library collection. College and Research Libraries, 44(4), 220–234. Kurth, W. H., & Grim, R. W. (1966). Moving a library. Scarecrow Press. Lambert, J. (2016, November 4). Confident shifting for complex moves [Lightning round]. Brick & Click: An academic library conference - 2016, Maryville, MO. https://brickclickanacademiclibrary2016.sched.com/event/79LV/lightninground-confident-shifting-for-complex-moves Lambert, J. (2022a). Evidence-based library book shifting [Spreadsheet]. Open Science Framework. https://doi.org/10.17605/OSF.IO/8M3CU Lambert, J. (2022b). Predictable book shifting. The Code4Lib Journal, 54. https://journal.code4lib.org/articles/16577 Leimkuhler, F. F. (1968). Systems analysis in libraries. American Society for Engineering Education, Lafayette, Indiana. https://files.eric.ed.gov/fulltext/ED047718.pdf Leimkuhler, F. F., & Cox, J. G. (1964). Compact book storage in libraries. Operations Research, 12(3), 419–427. Lindsay, B. D. (2017). Moving the New York University Abu Dhabi library. Collection Management, 42(1), 48–56. https://doi.org/10.1080/01462679.2016.1264900 Ravindran, A. (1970). On Compact Book Storage in Libraries. https://eric.ed.gov/?id=ED053781 Rider, F. (1949). Compact book storage: Some suggestions toward a new methodology for the shelving of less used research materials. Hadham Press. School of Operations Research and Information Engineering - Cornell Engineering. (2018). What is Operations Research? http://www.orie.cornell.edu/about/whatis.cfm Sheetz, A. C. (1932). The journey of the 360,000. Pennsylvania Library Notes, 13, 133– 136. Taylor, G. (2015, November 30). Moving Libraries and Archives: A Selected Bibliography [Digital Repository]. https://repository.si.edu/handle/10088/27701 Tucker, D. C. (1999). Library Relocations and Collection Shifts. Information Today. W., M. (1930). How the books were moved. The Yale University Library Gazette, 5(2), 30–34. JSTOR.