Determining losses in academic libraries and the benefits of theft detection systems

Abstract

The first part of the article reports on a study, undertaken at a small college library, to determine the accuracy of the statistical sampling techniques used for determining the number of books missing from library collections, com pared with taking a full inventory. A full inventory of circulating stock was performed involving a total of 5742 books (4000 general non fiction and 1742 specialist titles). The interval method of statistical sampling was used as a quick method of determining losses and five methods of sampling are described and discussed, those of: Bommer and Ford; Miller and Sorum; Clark; Niland and Kurth; and Griffith. Results for book losses determined via sampling methods were found to be very close to actual inventoried results, with a particularly acceptable result obtained when only one per cent of the collection was checked. The Bommer and Ford method came closest with a loss figure of 8.06 per cent compared with the inventory figure of 8.20 per cent. The second part of the article considers different ways of allocating a monetary value on replacing the lost material. The costs obtained (varying between 5 000 and £20 000) are compared and used as the basis for a cost benefit analysis of the introduction of library theft detection systems. Concludes that such secu rity systems would be justified, both in the short term or over a five year peri od. Provided the average price of replacing each book is greater than or equal to £13.56, the initial cost of the theft detection system and its first year's operating costs will be met by savings in replacement costs.