Quality-based model for resource allocation
Disciplinary domain level
The Consistory determined on a new model for quality-based resource allocation at their meeting in February 2018 (UFV 2016/1914, document in Swedish). The purpose of the quality-based resource allocation model is to provide clear signals in which dimensions the operation is to be strengthened. The resource allocation model is to stimulate quality development in research and postgraduate education. The starting point is that the model should provide good incentives and does not have reallocation as an end in itself.
In the quality-based resource allocation model, 7.5 per cent of the basic grant for research and postgraduate education is reallocated according to the following indicators, of which the second is a bibliometric indicator:
- External grants (2.5 per cent)
- Publication according to the Norwegian model and citations (2.5 per cent)
- Issued doctoral degrees (2.5 per cent)
The principle is that the outcome is calculated by comparing the disciplinary domains with their own achievements over time. The outcome of the last three calendar years is compared with the previous five.
The bibliometric indicator of the model
Publishing is an extremely essential part of research as it is through scientific articles and books that research results are communicated to both the research community and the rest of society. Publishing is also one of the components of the national model for resource allocation.
The bibliometric indicator consists of two components, each worth 50 per cent: the Norwegian model and the Mean Normalized Citation Score. For a description of the Norwegian model, see the Bibliometrics main page.
Mean Normalized Citation Score
Web of Science is used as the data source for this component. Only publications of the types “Article” or “Review” are included in the data. The data is retrieved from DiVA and transferred to CWTS Monitor, from where the normalized citation scores are retrieved for the years in question and for the respective unit of analysis. All publications are fractionalized according to the number of authors, even large multinational collaborations are fractionalized. Author self-citations do not count.
For both components, an average value is calculated for the outcome period (three years), which is compared with the average value for the comparison period (five years). The ratio between these is a change value, which is used to calculate the result from the indicator. Since citations can only be measured with reasonable reliability about two years after the time of publication, the outcome period is set to y-4 to y-2, where y is the year of analysis. For example, the period regarding citations for the analysis year 2021 is 2017-2019, while the corresponding comparison period is 2012-2016.
Detailed information on both the Norwegian model and the citation indicators can be found in the report Uppsala University Annual Bibliometric Monitoring 2021 (link to DiVA).