Availability analysis

A recently proposed technique for determining the availability of items in a circulating collection is expanded and applied. We present a discussion of the theoretical justification and a careful analysis of the sources of error and the reliability of the results. We report the results of a new longitudinal study using this technique and apply it retrospectively to a number of previous studies to which it is applicable. The resulting data indicate certain striking regularities in the performance of large university libraries, as well as significant differences. The implications of this tool, both for management and for the scientific analysis of various aspects of library performance, are discussed.