Weeding and Maintenance of Reference Collections

Linda S Katz

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781136579158
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Library & Information Science
• Index: Languages & Linguistics

Here is unique volume offering practical advice on weeding and maintaining reference collections. It covers different types of libraries--academic, corporate, public--and problems, and librarians describe in detail methods and criteria used by their libraries in weeding their reference collections. Dr. Pierce has organized the topics of her book into relevant chapters. These chapters, bound to appeal to a variety of needs, address and discuss the problems and management of growing reference collections. As many librarians find weeding reference books a difficult task, most reference departments suffer from a lack of space as a result. Collection growth reduces shelf and seating space, and both books and people are lost in the clutter. In reading this essential book, reference supervisors will come to understand the importance of allowing reference area growth combined with effective weeding to promote an attractive and well-stocked reference area. Heads of reference will find Weeding and Maintenance of Reference Collections full of useful information, from the specific criteria and detailed methods contributed by several librarians who have found success in weeding their reference collections, to the practical hints on planning and evaluating collection contents and organization. Students and faculty of library schools and information studies will gain insight into successful management of increasing amounts of reference material as the Information Age gathers momentum into the 1990s.

---

III. EVALUATING AND WEEDING COLLECTIONS

---

Discovering How Information Seekers Seek:
Methods of Measuring Reference Collection Use

Mary Biggs

SUMMARY. The author observes that reference collections are often too large for librarians and users to learn them well and fully exploit them. Evidence suggests that many titles in what were designed to be lean “working” reference collections are rarely if ever consulted. Systematically—athered use data are needed to guide selection and weeding decisions. Methods that have been employed to assess in-library materials use —including use of journals and circulating books as well as of reference works—are reviewed and their application to reference materials discussed. The author closes with concrete recommendations for carrying out reference collection use studies.

In an era when neither library budgets nor library buildings expand proportionately to the growth of publication, librarians need solid data to inform their collection decisions. Perhaps most needful and most neglected are reference librarians. Given their purpose, one would expect a library's reference works to comprise a lean “working” collection, well-known and thoroughly exploited by the reference staff. Additions to the collection would be made cautiously, and works that were rarely or never used would be weeded regularly for discarding or removal to another part of the library. Many reference librarians have, after all, remarked the critical need to “learn” the collection and the frustration of being defeated by its size.¹

In late 1985, Victor Biggs and I surveyed reference heads in 471 academic libraries of all sizes.² We discovered that even within libraries of comparable size and purpose, the numbers of titles held in reference collections varied greatly. In all of our four size strata, including small college libraries, collections numbering tens of thousands of titles could be found.³ And while all libraries weeded their reference collections, and substantial percentages indicated that “low use” by librarians or patrons was one weeding criterion, very few libraries had ever gathered use data systematically.⁴ This was all the more striking as respondents estimated that only about one-fifth of the titles in their reference collections were used by anyone as often as once a month (with some guessing as low as 1%); that slightly under half were used annually (with some guesses as low as 5%); and that only about two-thirds were used in a fiveyear period (with guesses as low as 10%).⁵

These guesstimates, however, were just that. And even if they were accurate, a reader of the article pointed out to us in a letter, they would offer little guidance. The obstacle to whittling the collection into a lean mean reference tool lies in the difficulty of determining which 80% is not used once a month, which half is not used in a year, and which 30% lies idle for five years and more. This observation circles back to the need for use studies. How can they be carried out, and what can they tell us?

Precisely because there have been so few, the library literature offers little direct advice. To discover pertinent models, I examined reports of research on in-library use of all types of material, most prominently of periodicals. Here I summarize the principal research methods employed, consider their applicability, and set forth suggestions for multi-faceted explorations of reference-collection use.

TOUCH TECHNIQUES

Perhaps the easiest to implement are what Fussier and Simon call “all or nothing” studies.⁶ Equating one use with many, these can at best yield only gross tabulations of works used, or moved, for any purpose in a given period. They are essentially of two types.

What can be termed the “Reach Out and Touch Technique” involves placing in each book (or in a selected sample of those books whose utility is doubted) some substance or item which will surely be disturbed if the book is moved. At the end of the study—which can extend over any period, though most likely a few months or a year-“disturbed” titles are noted. Among the substances that may be planted inconspicuously, Fussier and Simon suggest, are “infrared dust, beads on top of the book, or unexposed photographic paper inserted between the pages.”⁷ One portion of a periodical-use study at Newcastle University employed ordinary slips of paper.⁸

Obviously, the task of dusting, beading, or otherwise marking the books would be tedious and time-consuming, as would taking tallies to close the study (and sweeping up the beads before they tumbled readers to the floor). But a more substantive problem is the false registering as uses of mere physical disruptions. Books moved by shelvers to make space, or mistakenly pulled out by patrons misreading call numbers, or merely jostled by someone bumping against the stacks, may be recorded as “used.” The great advantage of this technique, however, is its independence of reader cooperation. “It has been found,” warns M. B. M. Campbell, “that the degree of success with any [use study] is inversely proportional to the level of user participation required by the methodology employed.”⁹

RESHELVING TECHNIQUES

Reshelving techniques, which may or may not fall into the all-ornothing category of research, require user participation but no user effort, and are probably the most popular methods because they can be integrated with normal library activity. Shelvers simply record items when putting them away. The variations are many, depending on the time available and just what the librarians wish to learn. At least in theory, all titles can be listed on a tally sheet ordered by call number or title, with space provided for the shelvers’ hatch marks. Alternatively, shelvers can list the titles of works as they are reshelved. But when tens of thousands of titles are involved, this will slow the shelving considerably. W. M. Shaw, Jr. suggests affixing a small pressure-sensitive label to a book's spine the first time it is reshelved¹⁰—which is an allor- nothing technique if the book is then ignored until the final tally. If the label is large enough to accommodate hatch marks, however, numbers of reshelvings can be recorded.¹¹1 used a related method when working at Bowling Green State University (Ohio). A small colored dot was pressed to the spine of each reference book the first time it was reshelved in a semester. This continued for two semesters and one summer, with dots of different colors used for each of the three time periods (i.e., fall semester was green, spring semester red, summer session silver). Although numbers of uses within a semester could not be told from this study, it did reveal whether the book was used consistently. When the study was completed, the subject specialists (all of whom also had reference and user education responsibilities) scanned the shelves in their call-number areas of expertise and noted those titles that had lain unused all year or had been used in only one semester. If a specialist wished to retain such a title in reference, he had to write a supporting rationale and have his request approved by the head of reference. A weeded title might be discarded, stored, put into circulation, or sent to the stacks as a non-circulating item—which guaranteed its availability in the building but reduced “noise” in the working reference collection. Obviously, a dot study could extend over a shorter or longer period: dots come in many colors.

Offsetting the simplicity and relative ease of reshelving studies are the fact that they cannot yield qualitative data and inevitably underestimate use. Everything depends upon used works being pulled off the shelf and left off. But unobtrusive observation and other checks on users’ habits show that even when signs ask them not to do so, many people reshelve the books they consult, especially when they have used them for only a few minutes. At Newcastle University, for example, twenty-one of fifty-five periodicals that were used were also reshelved.¹² Discussing a periodicals use study, Martin Gordon asserts: “It should be noted in passing that some feel… uses [marked by the need to reshelve] are equalled or even exceeded by [other] uses, but that [the former] are more indicative of need fulfillment.”¹³ Without supporting evidence, however, this is unduly optimistic and does not in any case deny the possibility of large distortions. To prevent patrons from reshelving, librarians customarily post signs asking them to leave materials at conveniently placed collection points, sometimes adding the explanation that a use study is underway.¹⁴ The study may also be publicized through talks to user groups, announcements in campus or town newspapers, and so forth—though Edward T. Shearin warns that too much fanfare may make readers self-conscious about their library behavior and alter normal use patterns.¹⁵

Figures may also be deflated when two or more people use a book before it can be reshelved, scooping it up from a table or collection bin. To minimize this problem, the authors of the American Library Association's influential manual entitled Output Measures for Public Libraries recommend gathering data only on sample days, during which books are reshelved punctually every hour on the hour.¹⁶

A problem with opposite effect is the fact that an unshelved work may have received no more than a cursory glance from its “user” or have fallen accidentally when a neighboring volume was pulled down—may, in any case, represent no serious use, no satisfied need.

In short, though the data gleaned by shelvers can be valuable, it will ideally be supplemented with other methods.

USER TALLIES AND SELF-ADMINISTERED QUESTIONNAIRES

Relying much more on users’ cooperation, yet a potentially rich source of qualitative data, is not-so-simply asking people what they use and how.

In a study of the periodical collection at the National Oceanic and Atmospheric Administration Library, Wenger and Childress attached to each current journal issue a slip of paper asking readers to make a check-mark if they use the issue “for any reason.”¹⁷ At Wolverhampton Polytechnic Library, Campbell placed punch-card sets in periodicals with the request that readers remove one card for each article consulted and place it (or them) in a handily located collection tray.¹⁸ Because such techniques require little of the user, they are thought (but not known) to elicit high cooperation.

More interesting and problematic is the method employed by Fussier and Simon in their famous study of research library book use. Eschewing simple counts, this team sought to answer four questions: why the reader had removed the volume from its shelf; where and for what purpose he would use it; and how valuable he expected to find it.¹⁹ In each sample book was placed a sheet of paper asking these questions. This questionnaire, which could not be seen unless the book was removed from the shelf, would be visibly disturbed if the book was opened —allowing the researchers to determine rate of nonresponse. That is, by counting questionnaires disturbed but not returned, they could tell how many people had used (or at least handled) a sample book but not filled out a questionnaire. In about half the sample books, cheap ballpoint pens were taped to the questionnaires, providing both an easy means of filling them out and a “reward.” Readers returned many more of the questionnaires that had been accompanied by pens, which, according to the authors, “hints at the degree of caprice in motivation” and, of course, gives still more information about nonresponse. That is, it can be assumed that “at least as many questionnaires as the difference between the two groups were seen but not returned.”²⁰ The disadvantage of the Fussier-Simon method is its complete reliance on reader cooperation, hence low response rate. However, it yields rich qualitative data. The questions asked, of course, would vary depending on the study's o...

Table of contents

• Cover
• Full Title
• Copyright
• Half Title
• Contents
• About the Editor
• Introduction
• I. CONSIDERING THE INSTITUTIONAL CONTEXT
• II. MAKING POLICY DECISIONS
• III. EVALUATING AND WEEDING COLLECTIONS