Weeding

I'm on a committee at work that is developing a plan to weed our library's collection of about 300,000 or so volumes. We need a systematic weeding program because we're running out of space. But besides opening up shelf space for new and more relevant library resources, a well weeded collection also can increase circulation. After all, with all the older, unusable things removed from the mix, it's easier to find the good stuff.

It goes without saying, but librarians don't like to throw books away. However, if a librarian takes a reasonable 20 minutes evaluating each volume (checking core lists, bibliographies, total copies in WorldCat, etc.), he or she will need at least 11.5 years. There must be a faster way.

There is! As a undergraduate liberal arts college, our primary weeding criterion can be use. Since my college is not a major research library, there isn't an expectation that we'll keep everything. As such, we librarians simply need to determine what books are being used, and withdraw the rest. And there will be evidence of such use on the date due slip or in the library system.

Also, using use as the primary criterion, we librarians avoid the problem that is encountered when using publication date to evaluate books: accidentally weeding those classics in a discipline published years ago. If a work is truly a classic, one would hope that faculty and students are using it!

We knew that we want to weed at least 10% of our collection, but we needed to determine a definition of "use" for our collection so removal at the shelves can be fast and efficient. So, we selected a statistically significant sample of 500 books (+/- 4.3% error, 95% confidence level). We recorded each book's last checkout date. We discovered that the bottom 10% of our collection last circulated before 1983. If the committee settles on the 10% figure, we'll be able to use this date as our test for each volume.

Yes, the system is not perfect: What about those books that were purchased 5 years ago but have not circulated? What about books in series, where one book was loaned, but none of the others? Or what about books that are used in-house and are never stamped with a due date? What about the books everyone is checking out, but are simply out of date?

Weeding is not perfect; accidents will be made. But a vast majority of things removed are never needed again. And more shelf space and an increase in general circulation are definite benefits.

We haven't started withdrawing anything yet. We just finished our sample and are trying to decide if we want to weed 5%, 10%, 15%, or all the way up to 30% of our collection. Whatever we decide, we have data to let us know when that bottom 5 or 20% last circulated, and at the same time have a convenient rule that staff can apply at the shelf to decrease the time agonizing over individual titles and increase the time we spend providing service at the reference desk, teaching in the classroom, or selecting new relevant information resources. Needless to say, we are excited to begin!

NOTE: For those of you who know where I work, our committee's work is in progress and our plan has yet to be approved by our library's faculty. Thus, the details described here are unofficial and are subject to change!