What about the Z collection? While it was a surprising revelation that 24 percent of the entire reference collection–the Zs–received only 2 percent of the total use, similar results have been reported by others. In Truett’s investigation of weeding policies and practices, she heard from “a couple of librarians” who mentioned the Z collection–particularly subject bibliographies–in their responses. One of the respondents’ practice was to put general bibliographies in reference, but to place subject bibliographies in the circulating collection, classified by its primary subject rather than Z. “Besides limiting collection size, it was felt this facilitates browsing by the general public, who are more likely to find a bibliography shelved with its subject.”26
Serendipity
The reference staff learned much that was unexpected during the course of this study.
The dotting process was very helpful when determining the disposition of gift books. The process for evaluating gift books always has included learning whether duplicate–or alternative editions–of the books are already held in the library. Due to limited space, reference staff avoided adding duplicate copies unless the books were heavily used. Until the dotting project was instituted, GAF had no way to determine objectively how much the existing books were used, and, therefore, the suitability of incorporating additional copies. Now, the bibliographer added to her procedure the practice of counting dots. This one added step tipped the scale toward more objective decision-making and was heartily embraced by the librarian responsible for gift books.
Until the preparatory work was done, the reference staff had only an estimate–albeit a fairly accurate one–of the number of reference collection books. Once the data were in the spreadsheet, the initial inventory completed, and corrections made to the spreadsheet, we had an exact number. The spreadsheet is being maintained, so the data are accurate and current. We developed methods to track acquisitions and withdrawn reference books separate from the records for the library as a whole, and identified and addressed years-old catalog errors.
The results of the study provided the means to evaluate the reference collection–as a collection and in relation to degrees provided by CIU. We have identified quantitatively which academic majors were well-supported, and where the reference collection was disproportionately distributed.
The Z collection became a surprise source of concern. As mentioned above, 24 percent of the books in the reference collection was in the Zs, an area that received only 2 percent of use.
We learned–almost too late–that it can be dangerous hiring intelligent student workers to do tedious, mind-numbing work. The students who were tasked with counting the dots and entering them in the PDA were excellent reference desk staff and good problem-solvers. Unfortunately, they sought shortcuts to the data collection process. One such shortcut involved entering zeroes in all the cells of the spreadsheet and replacing them as they encountered books with recorded use. Unfortunately, this failed to take into account items that were not on the shelves for whatever reason, and there were a substantial number of these. Between each change in classification, a row had been inserted in the spreadsheet. Here, too, the workers inserted zeroes. While the final number of zeroes was surprisingly high, it was far below what was first revealed when accessing the numbers entered by the creative, problem-solving student workers. As a corollary to this, we learned the maxim “Zero is a number, not a null value.”
Caveats
We are aware of at least three ways in which the data are flawed.
The GAF reference staff did not track acquisitions and deacquisitions as the project continued. For this reason, the statistics for each year reflect the same number of books, an apparent fact, but one that is most assuredly not true. When superseded editions were replaced, the dots were irregularly transferred to the new editions. Books that were weeded but not replaced were simply not accounted for in the system.