In the third year of the study, reference student workers used the reference shelf list in combination with the OPAC to create an electronic version of the shelf list. Over the next year, we reformatted these data and imported them into a spreadsheet workbook, with each sheet representing one shelving unit, and exported the workbook into a personal digital assistant (PDA). Next, we inventoried the reference collection by comparing the contents of the worksheet with the actual contents of the shelves. A wide range of discrepancies was identified (previously weeded volumes were still in the shelf list, volumes that should have been in reference were wrongly labeled for circulating stacks, shelf list cards for items that had been reclassified had not been updated, catalog records did not agree with the shelf list) and brought to the attention of technical services staff for correction.
At the end of the fifth academic year, three student workers began the process of counting dots. The students received print and oral instructions detailing every step expected, with a strong emphasis on strict attention to detail. The first student was to count dots and state the year and the number for the year. The second entered the data in the PDA, and then repeated back the information to the first student. The researcher admonished the students to read the numbers they reported, not simply repeat what they had heard or thought they had written. The third worker had the task of removing all but one dot per year from the books. The year and the number of dots for that year were written on the remaining dot.
We limited uploading of information to the reference librarian or the reference associate. Daily, they exported data from the PDA to the library’s shared network drive. From there, the reference librarian copied it to her PC and to a CD, which went home with her. As the summer of our discontent continued, reference staff agreed that with the learning curve, the problems with writing using the PDA, the screen size, memory, and speed, and the facility that most students now have with typing, the project would have been simpler using laptops instead. It was an interesting experiment, but not one to replicate if a laptop is available.
Findings
CIU opted to affix as many as ten dots per year, so the maximum a book might receive over five years was fifty. In so doing, we tracked use over time and more accurately assessed each book’s usefulness. Over the course of the study, we arbitrarily decided that books with twenty-one to fifty dots could be considered “heavily used,” those with six to twenty dots were “moderately used,” and items with one to five dots were “lightly used.” We found that 12 percent of the reference collection was heavily used, 17 percent was moderately used, and 36 percent was lightly used. While these data seemed positive, affirming the perceived–and actual–value of the reference collection to the GAF library patrons, there is one more number that must be considered: zero.
Fully 35 percent of the books surveyed in this study received no dots at all. Even in the most heavily used area–the BS classification, where the average number of dots per book exceeded twenty-two–fully 5 percent of the books on the shelves had no dots. While such a high number of zeroes was surprising, perhaps it should not have been. If the academic reference librarians who participated in the 1985 survey by Biggs and Biggs are representative, lack of use should have been predicted. Biggs and Biggs asked respondents what percentage of their collections they believed were used in four defined time periods and categorized responses based on institution classification. Responses were similar across the classifications: use increases over time, but even over five years, only 62 to 70 percent of the reference collections were estimated to be used.18 Empirical evidence seems to back up the intuition of the surveyed librarians, at least in their long-term predictions. Engeldinger’s report of the dotting project at UWEC’s library noted that 65.2 percent of the reference collection had been dotted at the end of five years.19 The Carlson Library at University of Toledo began a dotting project; Sendi’s findings–limited to only ready reference–indicated that after the first year, only 57 percent of materials had been used.20 Fishman and DelBaglivo, whose study excluded ready reference, reported a one-year reference collection use rate of 65.5 percent in the Health Science Library at University of Maryland.21 At GAF, we found that 64.7 percent books were dotted after five years. As with Engeldinger, CIU researchers found that more than 50 percent of the books in the reference collection were used fewer than twice in the five-year period of the study.22 Table 1 compares the estimates provided by the Biggs and Biggs survey with these studies.
In addition to simply measuring reference book use, the data were analyzed according to Library of Congress (LC) classification. In some ways, this aspect of the study revealed that reference staff instincts are accurate. As we could have predicted before the study started, books in the BR, BS (Christianity and the Bible) LCC ranges, as well as in the PA (Greek language and literature) and PN (literature) areas, received heavy use as defined above. At an institution in which biblical studies are preeminent, and in which the majority of students are undergraduates, these findings were by no means surprising. The high numbers in the Bs (and low ones in the Ts, for that matter) are misleading without the context provided by comparing the use by LC classification with the total population by LC classification. For example, the lowest non-zero number of uses occurred in the T classification (technology), where six books sported a meager eight dots. However, as the Ts hold only eighteen volumes, 33.3 percent of the books were used. While 33.3 percent does not represent enviable use statistics, it is not nearly as indicative of lack of use as the raw numbers would seem to indicate. Instead of focusing on depth–that is, the number of dots on a given book–we focused on breadth–that is, the percentage of books within a call number range that received any use. Eleven of the eighteen call number ranges in which the GAF reference collection holds books revealed 67 to 89 percent use. The remaining seven had 31 to 55 percent. We interpreted this to indicate that the librarians who made collection development decisions for the reference collection had effectively provided useful resources in all subject areas. In a context with a chronically inadequate budget, it was gratifying to learn that all areas of the collection were receiving use. Figure 1 shows numbers of books used versus those held by LC classification. Figure 2 provides the same data in terms of a percent.
Discussion
Use of Print Reference Materials
The reader might conclude that, based on the discussion above on zeroes, CIU reference collection use declined during the dotting project. On the contrary, during the five years in which the project was taking place, reshelving statistics indicate that use of print reference books increased by 40.4 percent. Consider the following factors that might have influenced this increase in use:
- increase in course-required use of reference resources;
- increase in library instructional sessions introducing and emphasizing reference resources; and
- increased reference librarian-initiated outreach to faculty yielding a concurrent increase in collegiality among the library staff and the teaching faculty.