Argues that a mix of quantitative and qualitative measures is most efficacious in library collection analysis. Cases drawn from the literature in which a combination of these measures was profitably applied are considered, and the types of software packages for purposes of data manipulation are discussed. (Contains 15 references.) (KRN)

Compares the more positive library automation experiences of a small public library with those of a large research library. Topics addressed include collection size; computer size and the need for outside control of a data processing center; staff size; selection process for hardware and software; and accountability. (LRW)