You are here:  

• Print

Responsible use of research metrics

Assessment of the quality and impact of research is part of the research activity lifecycle.

The University recognises the importance of assessing research responsibly and the value of peer review and expert judgment as a central component in the assessment of research outputs and wider research performance.

However, we recognise that quantitative metrics can have a supporting role in research assessment.

UoN UK Guiding Principles for Responsible use of Metrics in Research Assessment

 

1. Use Expert Judgement

• Research Assessment is based upon expert judgement and peer review. Quantitative measures can be used to support or inform judgements but should not be used to replace evaluation and critique provided through expert peer opinion.

2. Be Transparent

• Be clear what the rationale for using particular metrics is and how they will be used.
• Make data collection and analytical processes transparent as to whether, and how metrics will be used, including their calculation.
• Give researchers the opportunity to verify their data. This may involve releasing the data underpinning decisions, identifying data sources and metrics which will be used or asking researchers to source and submit data for review themselves.

3. Be Contextual

• The use of metrics should be appropriate to the research assessment being conducted and will vary according to the context e.g. institution-level metrics may be inappropriate to use at an individual-level. It should be clear from the outset of any research evaluation which metrics are being tracked and why.
• Value the intrinsic merit of the research output over the reputation of the publisher, the publication channel or associated metrics. This is a particularly important consideration if quantitative metrics are used as part of recruitment and promotion processes.

4. Be Fair

• Respect diversity of disciplines, career stages, employment pattern, gender and minoritised groups.
• Recognise that within scholarly communication (including peer review), systemic publication and citation biases exist which can disadvantage researchers from specific groups and consider ways to mitigate this.
• Variations between disciplines should be acknowledged. Use normalised or field-weighted citation data where possible and consider whether comparisons are meaningful.  
• Recognise there are a wide range of research outputs beyond journal articles, including monographs, conference papers, research data, preprints, working papers, protocols, and software.
• Recognise the range of contributions to 'team research'
• Recognise that some limitations in coverage of sources of publication data exists, e.g. unindexed publication venues and outputs written in languages other than English.

5. Be Robust

• All data should be accurate, and of an appropriate scale and scope, ensuring what is valued is measured rather than what is easy to measure.
• Choose a combination of metrics, this will ensure any weaknesses in one particular measure does not skew the data.
• Ensure the metrics and any associated evaluation methods are reviewed and updated regularly, recognising any unintended systemic effects of assessment and update to encourage appropriate behaviours.

University of Nottingham, Authored by UoN Libraries, Approved by the University Open Research Group (10 Oct 2024)