For all subject areas we can, at the very least, generate a measurement of the size of the collections broken down by classification ranges and/or some other variable. The numbers representing collection size can be useful as indicators of collection strength:

   (1) the relative strength of subfields within subject ranges,

   (2) the strength of subject areas relative to the library’s entire collection,

   (3) and in comparison with peer libraries.

Although it is possible to have a very large, poor quality collection, librarians tend to give size the benefit of doubt; in which case, the bigger, the better. However, in the absence of comparative data with other libraries, size can be difficult to interpret. Furthermore, size is increasingly tricky to establish. Print monographs, microforms, videos, and a few other format types are easily counted. E-books and e-journals in large serials packages, databases with partial full-text, and large e-collections not cataloged at the title level make counting collection size much more troublesome. Additionally, it is not uncommon for vendors to add and drop titles from large databases without notification to customers. The risks of under-representing the resources available to scholars are great. Hence, in circumstances of restricted time for assessment, when it comes to measuring strength, it may be best to limit the analysis to print monographs and/or to qualitative measurements such as the core lists or core collections described elsewhere in this work.

All caveats aside, local data can provide illuminating information. If the generation of comparative peer data is prohibitively time-consuming, relative strengths within a single collection may be addressed. Is a higher percentage of the collection found in just one call number sub-range or subject heading, e.g., 50% of the economics books are related to economic theory? Or, are the percentages almost the same across sub-ranges? In the case of the former, the collection could be described as "specialized," with its strength in theory; in the latter case, one might describe the collection as "generalized," with a somewhat even distribution of titles across subject headings. In either case, the results can be both interesting and useful.

For methods on obtaining comparisons with peer collections see techniques such as Public Catalogs and WorldCat and Gobi Peers.

Statistics on size might be used as indicators of collection strength and adequacy to answer questions such as:

Where do the library's total holdings rank among peers?

What are the strengths of the overall collection, analyzed according to subjects assigned to selectors?

Are there clear collection strengths within each subject area (does the collection has a narrow focus)? Or, are there some subjects wherein the distribution of the collection is equally proportioned across subfields (a broad perspective)?

How do these strengths compare with the same subjects within selected peer institutions?

When subjects are grouped into disciplines, in which is the collection weakest--Humanities, Sciences, or Social Sciences?

As a percentage of the libraries' total holdings, what are the top twenty-five subfields in the library? Are the bottom twenty-five collections too small and inadequate for the libraries' needs?

Compare collection priorities with selected peer libraries by sorting the data within subjects and ranking the subfield collections from highest to lowest percent: If "African Literature" is the strongest subfield within the library's "Literature" collections, how significant does it remain when compared to like-holdings in peer libraries? Among small collections that are low selection priorities or are weak, in which libraries do these appear as high priorities?

Data & Methods Bank