CWTS Paper in Nature
'Journal editorials give indication of driving science issues' written by Cathelijn Waaijer, Cornelis van Bochove and Nees Jan van Eck was published in Nature 463, 157 (14 January 2010).
- Two-level methodology
- The editorials of Nature and Science can be considered as an inventory of mainstream opinions in the international scientific community, and the choice of editorial subjects as a reflection of what the'community considers as important. We have analysed the choice of subjects from 2000 on, using a new two-level methodology:
- The first level is a mapping and viewing technique which groups the editorials on the basis of the co-occurrence of words into 15 clusters.
- The second level is a manual analysis of the content of a large sample of the editorials; this is used to verify the homogeneity of the 15 clusters and to clarify their subject.
- The results
- The results show a good correspondence between the two levels of the analysis. This means that the mapping and viewing technique is an efficient first approximation. The editorial subjects of Nature and Science are fairly similar, though there are intriguing differences:
- Nature pays more attention than Science to some very American subjects (NIH and NASA), and to research priority setting,
- whereas Science pays more attention to the impact of science on politics, and to issues related to developing countries.
