|
Why do you ask? 4. The purpose of reference statisticsThe cost is too high and the benefit too uncertain for libraries to initiate massive data collection on all activities and services. We already have decent data on loans and on library visits. Reference is a natural third. Inquiries emerge from the world of the user - but are processed in the world of the librarian. In order to manage the flow of questions we need good categories and sorting procedures inside the libraries. In order to satisfy the users, we need much more information on the queries themselves, on the user profiles, and on their use of the answers. In the field of lending we are swamped by data. In the field of reference we are starved. Our public library statistics give information on loans to adult and loans to children. We distinguish between fiction and non-fiction. We provide separate data on music, on spoken phonograms, on videos and on CD-roms. We lump, I admit, musical notes, photographs, microforms, multimedia sets, graphics and dias series into one category. But the total is counted. And we do this every year for every single municipal library in Norway. Table E. Lending and reference statistics. Norway 2000.
Source: Statens bibliotektilsyn (Norwegian Directorate for Public Libraries) When it comes to reference, the situation is very different. The number of inquiries is, in principle, collected by all public libraries during one typical week in mid-spring and one in mid-autumn. Multiplied by 26 this should give the annual volume of questions. The national public library authority provides a one-page guide on counting procedures. But they suspect (rightly) that collection standards are highly variable - and thus do not publish the data in printed form. Starting with the year 2000 the data are, however, available on the web. Reference statistics constitute a poorly developed area of library statistics. Usually, only the number of queries have been counted. Our knowledge about the core of reference work - the actual content of questions and answers - is very limited. Today, however, the web is transforming reference work. Web technology is reducing the number of questions addressed to libraries. Academic libraries in the United States, which have the most complete time series, report a decline of more than 30% between 1997 and 2001 (ARL). Users are becoming more self-reliant. But e-mail and databases are also used to establish virtual reference services that store, index and reuse transactions. This creates new possibilities for measuring, studying and improving reference work. We collect statistics in order to compare. We want data that reveal changes between years, differences between libraries and distinctions between countries. But comparative statistics are only meaningful when they are based on shared and meaningful concepts. If statistics of library visits cover both physical and virtual visitors, it is impossible to know what changes in the aggregate mean. If one library counts all queries, say, and another excludes directional questions, we can not compare them. Pprofound agreement is not necessary, however. In order to get usable statistics, three things are needed:
|