UCL has recently launched its new bibliometrics policy, which sets out principles for the use of citation metrics in research assessment across the university. It aims for sensible, fair, and balanced use of metrics in research assessment that values research and researchers on their own merits, moves away from some of the more inappropriate methods like focusing on the impact factor of journals or the h-index of authors, recognises diversity in research practice and outputs, and emphasises that the use of citation metrics is not mandatory.

This is an important step in supporting the use of Open Science and Scholarship across UCL. A key aspect of the open science movement has been in challenging traditional ways of disseminating research – whether that be through publishing in Open Access journals, opening up peer review, disseminating work at an early stage via preprints, or a range of other methods.

Many of these approaches, however, do not fit well with traditional methods of assigning credit using citation metrics.

For example, a relentless focus on the impact factor was a barrier to early adoption of open access journals. Newly created Open Access journals – which did not qualify for an impact factor – were seen as lower quality than the established journals, deterring authors from submitting to them. Similarly, megajournals, which did not cherry-pick papers for “significance”, had impact factors substantially lower than more selective titles – an author who was being judged on impact factors would be less keen to publish there.

In addition, limitations of the citation databases can penalise supporting material like data or code, which are often not indexed properly – if they are cited at all. This makes them appear less significant than they are. Similarly, preprints often get the majority of their citations before they are “published” – and these may not be tracked or credited accurately.

Factors like this mean that a focus on using traditional metrics can actively deter people from adopting Open Science approaches for their articles or their data. It is of vital importance that the ways we assess research do not discourage people from being able to conduct their research in the way that is best for them, and best for the wider research community.

Our new policy tries to move away from traditional uses of metrics, emphasising that citation-based metrics are not always appropriate and we do not have to use them if they’re not generally accepted in the field. Where they are used, we should avoid trying to impose a one-size-fits-all model, and consider all works in context.

Alongside the policy, we have provided detailed guidance for using alternative metrics, going beyond the impact factor or simple citation counts to assess citations in the context of other comparable work. We have also created the video below, and a Moodle module to walk you through the key elements.