View article

There is an increasing demand for science to help in addressing grand challenges or societal problems, such as tackling obesity, climate change or pandemics. In this context, it becomes important to understand what different sciences can offer to tackle these problems, and towards which directions scientific research should be developed. A useful starting point is to investigate what is the existing science supply, and which research options are better aligned to address grand challenges and societal demands (Sarewitz & Pielke, 2007). In order to map the science supply, we need a representation of the knowledge on research topics relevant for a problem.

Bibliometrics can provide very helpful tools for developing knowledge representations. However, these representations are highly dependent on the data and methods used. As a result, bibliometric tools or indicators often reproduce the biases in the data collection and treatment. For example, it has been shown that conventional bibliometric analyses are biased against non-English languages (Van Leeuwen et al., 2001), developing countries (Velho & Krige, 1986), applied science (Van Eck et al., 2013), the social sciences and humanities (Martin et al., 2010) and interdisciplinary research (Rafols et al., 2012). The aim of this paper is to investigate the biases introduced by available databases in the representation of research topics. In a previous study on rice research, we showed that the bibliographic database CAB Abstracts (CABI)–which is focussed on agriculture and global health–has a larger coverage of rice research for most low income countries than Web of Science (WoS) or Scopus …