Research networks are social spaces in academia where knowledge processes happen, driven by collaboration and competition forces. The diversity of different networks spaces offers a broad spectrum of understanding, interpretations, and operationalization of research networks. We can find some relevant articles within research networks scope. Therefore, we decided to classify those studies into micro-, meso-, and macro-scale dimension (Dopfer et al., 2004; He et al., 2011). Therefore, some examples of articles are given on that spectrum, from studies focused on individual researcher’s networks to global and knowledge networks. In our studies, we chose to use three dimensions of analysis (micro/meso/ macro), which range from individual (as ego networks) to global (as knowledge network) levels. Micro meta-level includes researcher networks and project research and group research. We considered organizational and institutional levels belonging to meso meta-level. At Macro meta-level contains National research system, international and global research systems, and disciplines or scientific fields levels. This sequence is hierarchical but only on scale issue, that is, no level is more important than the other. For each context in the study, we should select and define the most appropriate level to address the research questions. Sometimes, it is possible to use more than one level, that is, one might examine the impact of science politics and related programs, macro-level, which promote formal networks, micro-level on institutional or meso-level.
Note that this system of levels analysis classifies research networks on three dimensions: micro networks, meso networks, and macro networks and its subcategories. Other classifications can help the analysis networks: formal versus informal networks; short-term versus long-term duration networks; highly bounded versus more fluid networks; simple versus complex networks; internal and external networks; and international versus domestic (Allen et al., 2007; Glanzel et al., 2006; Glanzel and Schubert, 2005; Helble and Chong, 2004; Lemarchand, 2012; Leta et al., 2006; Lowrie and McKnight, 2004; Newman, 2003).