The Cost of Research

Abstract: Tools of empirical social research (data collection and analytical methods) contribute to environmental pressures in a similar way as transport, consumption or livestock production. In the latter context, we are intensively investigating the extent of their ecological footprint and the various options for reducing it. In contrast, the choice of tools in social science practice is currently determined primarily by the cost of research, the time and human resources required. In addition to collected data, there is also a huge amount of transactional or even social media data, where the computational capacity of complex models can be an additional environmental burden in the analysis. Determining the ecological footprint of social science research can be a key step in developing sustainable scientific practice. In this article we review the international literature on this topic, highlighting existing good practices. In the analysis, we present the example of NLP and show that a more conscious consideration and reduction of the ecological footprint of social science research can contribute to the development of sustainable scientific practice. Our analysis helps to develop an approach that can reduce the environmental burden of research without compromising the quality and reliability of scientific results.