![Sample our Education journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5932/TOC-Subject-Sample-2021-Education.jpg"})
Abstract
This article investigates the extent to which the Danish Bibliometric Research Indicator (BRI) reflects the performance of men and women differently. The model is based on a differentiated counting of peer-reviewed publications, awarding three and eight points for contributions to ‘well-regarded’ and highly selective journals and book publishers, and 1 and 5 points for equivalent scientific contributions via ‘normal level’ channels. On the basis of bibliometric data, the study shows that the BRI considerably widens the existing gender gap in researcher performance, since men on average receive more BRI points for their publications than women. The article suggests two probable explanations: (A) women merely comprise 24% of the committee members determining which publication channels to classify as ‘well-regarded’ and ‘normal’, which may lead to biases in the classification process. (B) The model privileges collaborative research, which disadvantages women due to gender differences in collaborative network relations.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1. The current research funding system operates on the basis of a two-tier model consisting of basic research grants allocated directly to the universities and external funding allocated via the national research councils, the European funding programmes and others (Frölich, Schmidt, and Rosa Citation2010).
2. Already prior to its inauguration, Danish academics were raising concerns that the BRI would also be used for other purposes, for example to incentivise the publication behaviour at lower management levels and eventually to hire and fire staff (Wright Citation2014). The Danish BRI is developed on the basis of the Norwegian Research indicator (for an introduction to this model, see Schneider Citation2009).
3. Danish Universities is a stakeholder organisation ensuring and maintaining the collaborative ties between the universities. The organisation initially opposed the introduction of the BRI (Jensen Citation2011).
4. The globalisation pool was initiated in 2006, in the wake of the so-called Barcelona Objective, which called EU member states to dedicate 3% of their GDP to research by 2010. The pool, which extended through 2012, substantially increased the budgets of the Danish universities with the aim of making Denmark an internationally leading knowledge economy (Wright Citation2014). As one of a few European countries, Denmark achieved the Barcelona objective – a year ahead of schedule – with a GDP expenditure of 3.09% dedicated to research in 2009. Approximately two-thirds of these resources were invested in the private sector, which is characterised by a large pharmaceutical industry (DMS Citation2011).
5. Since 2013, the BRI has been applied to an earmarked share of the overall basic funding, varying from year to year. This so-called ‘conversion pool’ still allocates a minor share of the overall basic funding.
7. The Rector College is the decision-making organ of the stakeholder organisation Danish Universities. The college consists of the eight Danish university rectors.
8. Committee compositions are not static, and at least some of the committee members are substituted on an annual basis.
9. This framework was replaced by the Research Excellence Framework in 2014.
10. Basically, the GLM connects the data to a linear predictor on the basis of a link function. As pointed out by Wu (Citation2005, 1031), this link function helps ‘accommodate a variety of nonlinear response distributions and map the linear predictor onto the range of these distributions’. Since one of the dependent variables (the differentiated count of BRI points) is based on positively skewed non-integer values, the GLM gamma error distribution and log-link function is considered the best-fitting approach for this study (see McCullagh and Nelder Citation1989).
11. As illustrated in (appendix), the female share of associate and full professors in the natural sciences and technology in Danish academia is particularly low, which serves to explain the modest samples sizes for this group in the data set.
12. In this regard, it should be noted that the Norwegian research and university system shares many similarities with the Danish system.
13. Existing data reveal that women comprised 26% of the committee members in 2008 (i.e. the first year of the BRI model) and 26% in 2015 (DMS Citation2015b; Henningsen Citation2008).
14. It is relevant to note that this result is not an effect of horizontal segregations (i.e. more women tend to do research within the humanities and social scientific fields than within the health and natural sciences). A statistical analysis of the distributions among health and natural scientists confirms the overall pattern on this matter.
15. Espeland and Sauder define reactivity as a process wherein ‘individuals alter their behaviour in reaction to being evaluated, observed or measured’ (Citation2007, 6).