We analyze the benefits in terms of scientific impact deriving from international collaboration, examining both those for a country when it collaborates and also those for the other countries when they are collaborating with the former. The data show the more countries there are involved in the collaboration, the greater the gain in impact. Contrary to what we expected, the scientific impact of a country does not significantly influence the benefit it derives from collaboration, but does seem to positively influence the benefit obtained by the other countries collaborating with it. Although there was a weak correlation between these two classes of benefit, the countries with the highest impact were clear outliers from this correlation, tending to provide proportionally more benefit to their collaborating countries than they themselves obtained. Two surprising findings were the null benefit resulting from collaboration with Iran, and the small benefit resulting from collaboration with the United States despite its high impact.

The present paper proposes a method for detecting, identifying and visualizing research groups. The data used refer to nine Carlos III University of Madrid departments, while the findings for the Communication Technologies Department illustrate the method. Structural analysis was used to generate co-authorship networks. Research groups were identified on the basis of factorial analysis of the raw data matrix and similarities in the choice of co-authors. The resulting networks distinguished the researchers participating in the intra-departmental network from those not involved and identified the existing research groups. Fields of research were characterized by the Journal of Citation Report subject category assigned to the bibliographic references cited in the papers written by the authorfactors. The results, i.e., the graphic displays of the structures of the socio-centric and coauthorship networks and the strategies underlying collaboration among researchers, were later discussed with the members of the departments analyzed. The paper constitutes a starting point for understanding and characterizing networking within research institutions.

Purpose -The purpose of this paper is to visualize the inter-university and international collaboration networks generated by Spanish universities based on the co-authorship of scientific articles. Design/methodology/approach -The approach takes the form of formulation based on a bibliometric analysis of Spanish university production from 2000 to 2004 as contained in Web of Science databases, applying social network visualization techniques. The co-authorship data used were extracted with the total counting method from a database containing 100,710 papers. Findings -Spanish inter-university collaboration patterns appear to be influenced by both geographic proximity and administrative and political affiliation. Inter-regional co-authorship encompasses regional sub-networks whose spatial scope conforms rather closely with Spanish geopolitical divisions. Papers involving international collaboration are written primarily with European Union and North and Latin American researchers. Greater visibility is attained with international co-authorship than with any other type of collaboration studied. Research limitations/implications -Impact was measured in terms of journals rather than each individual paper. The co-authorship data were taken from the Web of Knowledge and were not compared with data from other databases. Practical implications -The data obtained in the paper may provide guidance for public policy makers seeking to enhance and intensify the internationalization of scientific production in Spanish universities. Originality/value -The Spanish university system is in the midst of profound structural change. This is the first paper to describe Spanish university collaboration networks using social network visualization techniques, covering an area not previously addressed.

No abstract