What's the use of bibliometrics ?

Abstract

Bibliometrics has been heralded as science. But do the mathematical the theoretical base of information models which it develops have any practical value ? A number of applications have been proposed; few have actually been implemented. Some are based on flimsy evidence, some are hyper-complicated, some are simply impractical. This paper develops a couple of scenarios in which bibliometric theory could have a practical value. The particular application areas are file design and journal collection management. In the examples presented here, procedures are introduced for estimating the parameters of the Zipf / Lotka and Bradford distributions from samples. However, if these applications are to be of maximal use, more information about the errors of estimate should be included. This is an area to which bibliometricians could usefully turn their attention.