Bibliometrics

Bibliometrics at the University Library

Our bibliometrics service can, essentially, be divided into two parts: bibliometrics information and bibliometrics analyses.

Information ranges from participating in seminars where we discuss the meaning of bibliometrics, the types of analyses that can be conducted, how bibliometrics is used for evaluation and its research policy context, to giving advice to researchers who wish to know how to approach requests to specify the h-index in applications for research funds. Furthermore, we also contribute with comments on external analyses of research at Lund University as well as advice in relation to the acquisition and use of tools for bibliometric analyses.

Analysis comprises everything from publication and citation analyses as a basis for decisions concerning the allocation of funds on behalf of, for example, faculty management, through additional analyses of external evaluations of research projects connected to Lund University, to surveys of research collaborations commissioned by research groups prior to evaluation or as part of the preparation for collaboration with, or visits to, other higher education institutions.

Bibliometrics and evaluation of research

Bibliometrics is increasingly used for the evaluation of research and the allocation of research funds. Many of the international university rankings that have attracted attention in recent years are at least partly built on the use of bibliometric indicators.

At Lund University, research funds for the Faculty of Social Sciences and the School of Economics and Management, for example, are allocated on the basis of different kinds of publication analyses. Also, research funders increasingly demand bibliometric statistics from researchers in their announcements of research grants.

• Field-normalised citation indicators are used to relate citations to local publication to scientific publications globally, normalized for research fields and publication year; for instance the share of local publications among the 10% most cited publications globally.
• Journal Impact Factor measures the average number of citations for articles in a journal.
• h-index indicates the number of publications cited a specific number of times.
• In Norway, funds are allocated according to the number of publications, taking the type of publication into account – books, anthology chapters, and journal articles are awarded different “pointsˮ – but also the status of the publishing house or journal in which the text was published (the so-called “Norwegian listˮ).

Calculations can also be based on many different data sources: in the Scopus- and Web of Science databases, journal articles are indexed along with their references, which means that citation calculations can be made. At LU, we have access to SciVal, a platform for bibliometric analyses based on Scopus data, through which a number of various bibliometric analyses can be performed, for instance citation analyses of different kinds. Publication databases such as LUCRIS can be used to make calculations of different types of publications, and even Google Scholar can be used to retrieve statistics on how many times a publication or an author has been cited. In the case of Google Scholar, you have to be extra careful in interpreting the figures since the quality of Google's citation data may vary.

• Web of Science
• Scopus
• SciVal
• Google Scholar
• LUCRIS

However, bibliometrics is not only about evaluating research by measuring productivity reflected in the number of publications and the impact shown by the number of citations. Depending on the availability of data, but also on how different research fields are organised, bibliometrics can be used to analyse different structures within research and research fields. For example, through information on authors’ addresses, research collaboration can be mapped at the individual, organisational and national level. Based on how authors or journals occur together in reference lists - but also analyses of terms used in scientific text - different research specialisations can be mapped within a field, for instance for an overview in a literature review.

The different methods measure different aspects of publications and citations. They vary in degree of complexity and requirements on available data and there are also vast differences in the extent to which an indicator may be suitable for comparing publications from different research fields, for example, or whether it is at all useful for analysing publications from a particular field. Furthermore, there are vast differences between different data sources, both in terms of content and the quality of how the content is described and analysed.

Please contact the University Library if you have any questions regarding bibliometrics.