Descriptive Versus Evaluative Bibliometrics

Abstract

This paper covers the differences between two separate bibliometric approaches, labelled ‘descriptive’ versus ‘evaluative’, or top down versus bottom up. The most important difference between these two approaches is found in the level of validity of the underlying research output. Whilst the publications in a top down approach, having a descriptive character, are collected by following general characteristics of these publications (such as country names, or fields), the consequence is that findings from such studies have a ‘meaning’ that is limited with respect to actual research assessment. On the other hand, in a bottom up approach the publications are collected from individual oeuvres of scientists, including a process of verification by the researchers involved. This procedure contributes significantly to the validity of the publication material, and consequently research assessment procedures can be based on the results of this type of bibliometric analyses. A strong focus in the paper will be on the actual application of bibliometric analysis within research assessment procedures, in particular within the UK and the Netherlands.