Column Editors: Stacey Marien (Acquisitions Librarian, American University Library)
and Alayne Mundt (Resource Description Librarian, American University Library)
Column Editor’s Note: In this month’s column, we feature the experience of relocating volumes off-site using publication date and circulation data as selection criteria for relocation. Matthew Smith, Manager of Circulation Services at American University Library, explains the criteria and process for moving a large number of titles to our consortium’s (WRLC) off-site storage. — SM & AM
For the second time in five years, American University Library is moving a sizeable portion of its general collection to a shared off-site storage facility. With planning already underway for a modest library renovation, it is being done to rebalance space between collections to create a more dynamic and collaborative environment for our university community. At this point, I would like to take an opportunity for reflection on the paths that we’ve taken and to compare how this second move of fifty thousand volumes will differ from the first move of one hundred thousand volumes.
Before getting too technical about the process itself, let’s consider the rationales for this method of using publication dates and circulation data, over a more hands-on selection process. As creating space is the primary reason for the move, time is the motivation for the method. The amount of time required for subject selectors to make decisions on a title-by-title basis, along with the necessity that these decisions be made quickly so as to not compromise the timeline of the project, makes an automated selection process appealing. Furthermore, such a data driven and neutral system can be applied uniformly to the whole collection.
Selecting Criteria and Pre-Work
First of all, before a project like this is started, there is probably a specific amount of space that is desired, be it an entire floor, a certain set of ranges, or to generally make space for collections. For the removal of shelving, one must first determine how many books are accommodated in this area and ensure that it is a healthy amount relative to the distribution of the collection in general (i.e., one would need to move more books than contained in this area if other areas are cumulatively overcrowded).
In undertaking such a project, proper space distribution throughout the collection and accurate inventory of the collection to be handled will produce better results; the number of books needing to be moved will be better determined and reporting will be more effective at forecasting appropriate criteria and will more accurately yield the correct number of books for relocation.
In our experience, we have used two basic points of consideration for selecting books for relocating: circulation and publication dates. Circulation has been based on the last time an item was borrowed by a patron and publication date is, as expected, the year the book was published; the date of bibliographic record creation is also checked, so as to avoid sending out anything that was recently added to the collection for the first time.
I think it is best to think of the selection variables as dials that one can turn up or down to arrive at criteria that results in the correct number of books for relocation. As examples, our first project to remove 100,000 items was based on a publication date of 1980 and earlier with no circulation after 1997, the year that our library migrated to our current ILS, Voyager. Our next project will use a publication date of 1990 and no circulation since 2000 to identify 50,000 books for relocation to offsite stacks.
Reporting is key to this process, and it is probably fair to say that access to data can differ dramatically from one library to another for a variety of reasons. Here, our process has benefited by direct access to Voyager circulation and bibliographic data through Open Database Connectivity (ODBC) via Microsoft Access. In short, I was able to identify all records by publication date that met the circulation criteria and construct reports that had all the necessary information. Due to the large amount of data, it was frequently necessary to break up the work, going back and forth between Access and Excel to ultimately arrive at a list of items meeting the criteria and to be pulled. In this current project, I paid especial attention to the distinction between monographic and serial titles.
Types of Books and Special Criteria for Multivolume Sets
This strategy is straightforward for monographs; however, multivolume sets and the analyzed series can be a challenge. First, a few general points: For catalog consistency, we have always taken an all-or-nothing approach to multivolume sets; i.e., criteria that determines whether an entire set either stays in the building or goes off-site, so that sets are not split up between locations. The first move to storage project utilized an 80% threshold for off-site relocation, requiring that the set would need to meet both circulation and publication dates criteria by at least 80%, or the set would remain on-site. In our new project, we have adopted a threshold of at least 50%, so that a set can go off-site as long as at least 50% of the items meet both the circulation and publication dates criteria.
Multivolume sets, particularly older ones, can present a challenge in that all volumes were not always given item records in the past. Here we discovered a side benefit. Sets that were relocated received some much needed attention — as items were created, sequencing was corrected and other cataloging enhancements and updates were made.
It is here that the experience of the first move was particularly informative toward process improvements for the second. For the 100,000 volume move, not much attention was paid to isolating the multivolume sets; volumes of a set were kept together, though the sets were intermingled with monographs in the staging area. In our new process, not only are sets isolated, they have been subdivided into two categories based on reporting. The first would be, for lack of a better term, “normal”; i.e., these sets are typically cataloged as items under a single bibliographic and holdings record. The second category contains sets that may present more of a cataloging challenge; these include items that contain several titles bound together in one volume and analyzed volumes with unique bibliographic and holdings records. These were also tested by the multivolume set criteria in this phase; whereas, in the first project, that work was done during the pull. In general, more work was accomplished in the reporting phase in an effort to make the pull more seamless.
The Pull Itself
Pulling of the books has been handled primarily by student staff. If possible, shelf-reading in advance makes the process go smoother. Much of this work can be done independently by the student workers after basic training; however, guidance with multivolume sets can be necessary for the longer term, especially when a set contains multiple bibliographic records where reporting and proactively applying the criteria can be a challenge. During this process, it is important to double-check that produced reports line up with what is physically on the shelf, keeping up a kind of a reality check. During the first project, there was a much higher number of records without items and it was important not to mistake another book on the shelf (with the same title or call number) for one of those on the list without an item record. It is also an important opportunity to confirm that publication dates are as expected and definitely meet your criteria. After books are pulled, we take a moment to vacuum and count them prior to shelving them in a staging area. At that point, depending on the scope of work needing to be done, either student or full-time staff in the Cataloging unit completes the work necessary to prepare the items for relocation to the off-site collections facility.
Prior to the first project, there was some concern that patrons would not react positively to having so many items moved off-site. Years out from that move, there has not been any measurable negative reaction. This is most likely due to the low circulation and age of the relocated books. It is also worth noting that many of our patrons request to have their books pulled for them via the catalog and that these items, now located off-site, are deliverable in one to two days to our main library, so the requesting process and experience is much the same regardless of a book’s location. From a browsing perspective, the argument could be made that the collection becomes more up-to-date and relevant, when older, low use titles are removed. Also, as with the multivolume sets, many monographs received item corrections and cataloging enhancements that made them more discoverable and generally in better shape after being processed for storage.
Our first project of sending out 100,000 volumes resulted in the addition of 70 spacious, individual study spaces, while our next move of 50,000 volumes will enable us to clear half of our lower level, by consolidating our general collection on the second and third floors of our library.