Making Collection Development Friendlier: Wishful Thinking

Collection development has been an saga for me perhaps in part because I never took a collection development class. (Drexel University, where I got my Master of Science degree in Library and Information Science, doesn't require it.)

I was the first subject librarian hired at Arcadia University almost 4 years ago. I am subject liaison to the general and special science departments on campus and in that capacity conduct librarian research instruction sessions. I also provide collection development support to those departments.

Over the last 4 years, I've tackled CD on a task-specific basis. In Arcadia's Bette E. Landman Library, we've completed numerous projects that have helped me develop the way I think about our collection. We've evaluated our print serials collection. We've replaced missing books. We've weeded our collection. We're constantly recommending books to our departments to purchase. We process requests for new journals. And I'm still working on the way I see collection development.

Knowing myself, I'm not surprised at this learning curve. I've learned over the years of being a professional in several different fields that it usually takes about 3 years for me (for many people, in fact) to become comfortable with the responsibilities of a particular job. Since I don't do collection development as a primary function in my job--plus because there are so many facets of it--I'm still getting comfortable after 4 years.

With the end of the fiscal year approaching, June 30, our current effort is get departments to spend their department allocations. As a library, we have our budget divided among several line items. The book budget gets allocated among all the university's departments to ensure a well-balanced collection.

What I've discovered about our book collection because of the various projects we've done for it is that it seems to represent a kind of bell curve across time. (This is a guess, because I've never quantified it. This observation is also limited to our sciences--LCCO: Q.) Our oldest books seem to be 80-100 years old. The peak of the curve seems to be 1980-1990. Then the tail seems to drop off. Obviously we need to purchase more heavily to make our collection more current.

But if the collection should meet the needs of our users across the disciplines and the course subjects, the task of simply buying more books is complicated.

Here's what I need to help me purchase well for our collection. I need a holdings-based application that can analyze our collection.

1) First, I want to be able to see the course topics laid out across the LC call number range as our collection currently looks. Makes sense, right? Our collection should serve our curricular needs at some level. Perhaps even better, though much more challenging, is to map course topics across LC subject headings; challenging because books can only have one call number in a given collection, but they can have many subject headings. One book may serve multiple courses. But such a book doesn't contribution to a very specialized collection.

2) Since our collection peaks in numbers of books around the 70s and 80s, I want to see a representation of topics across time.

3) I also want to be able to superimpose the idea of priority. We may not want to develop historical topics in our collection, but we may want to purchase in newer or more specialized areas.

4) Let's make this application more robust. I need to see our collection weighted by the number of students registered in the courses served by a given topic (as identified either by call number range or subject heading). Courses with large enrollments using a large number of books should have the same relative number of books serving them as courses with small enrollments using a proportionally smaller number of books.

I need to see all these things in a way that I can manipulate and modify so that I can see where I need to be recommending purchases.

That's not asking too much is it?

OCLC actually has a collection analysis product that can do some of what I want. Because Landman Library is a member of OCLC, it can look at our holdings and compare it to the holdings of all its member libraries, including ones of comparable size. It can analyze by publication date. It costs $500 to set up the service. The annual fee is calculated separately. But OCLC does not have access to our enrollment information. I don't know what visualization capabilities it has. I just registered to demo it, so I can find out.

I'd love to see if I can find an information sciences student at Drexel University's iSchool who could analyze this problem and imagine a solution, if not even design one. Dr. Chaomei Chen has done considerable work on information visualization, although not quite in this area. While a student there just over 4 years ago, I went to a presentation of his, so I know of his work.

Again, am I asking too much?