Most of the work done in development is done in collaboration, in partnership with individuals or organizations who contribute to a particular task or project we are working on. These collaborations are sometimes very straight forward, but sometimes they are quite complex, and involve many links and relationships.
With that in mind, I would like to share an approach I am working on, Social Network Analysis (SNA). We are using SNA to study research networks, its characteristics and how the network contributes to better research outcomes.
Social Network Analysis, or SNA, is an approach that studies social relationships within a network. SNA identifies which individuals are best at passing information or which have the most influence. It also looks at how individuals are connected (or not) to one another through the network (Newman, 2003), in other words, the types of relationships and ties they establish.
Social networks have been studied and analyzed for almost a century, but with the advances of statistical models and computer software, they can be analyzed and presented in a more sophisticated way, using a much larger quantity of data. It is also important to acknowledge that social networks have evolved, from much simpler and smaller networks to much larger and complex ones, requiring different tools for analyzing them.
SNA is applied to many fields, such as a research tool on human behavior, or as a tool for monitoring and evaluating a project or program, to name a few. It uses quantitative and qualitative data
A good example of how SNA is used as part of research in the health sector (in this case HIV-AIDS) can be seen in a video presentation by Eric Rice from the University of California (UCLA). Here, the study looks at sexual behavior and how it affects the spread and prevention of the disease.
YouTube source: Eric Rice, University of California (UCLA) [1 hour 10 min]
SNA can complement other existing approaches or methods for M&E, such as the logical framework where, instead of a linear model (activity-output-outcome-impact) describing events, it can describe actors and their relationships. The figure below, from Davies (2009), illustrates the SNA and a logical framework, side by side.
Image source: Davies, R. (2009, p.9) http://betterevaluation.org/resources/overview/Davies_Use_SNA
Collecting data and drawing maps related to the network can be performed by an outsider, such as a consultant, but interpreting the information or navigating through the network, should be done together with those involved in the project, who know the linkages, the relationships, the context and historical facts that contributed to the current structure of the network. With this dialogue and interaction, the SNA would have more meaning to the network itself, and can contribute to future decisions on the network activities.
Ways ILAC has used SNA
The ILAC Initiative has used SNA on two different occasions. The first project aimed to position a research organization within a larger network of actors. The exercise of analyzing the organization’s network was particularly useful as it provided an opportunity for staff members collectively, to identify the stakeholders they were related to, and those who were missing from the network, as well as the powers of influence and why. It also provided an opportunity for the research organization management team to listen to some of the influential actors identified in the network, learn about the role they play (or should play) in the network. The SNA, in this case, was used as a strategic tool for research priority setting, strengthening partnerships, communication and fund-raising.
The second example was a recent project where ILAC used the SNA approach to develop a system for monitoring the evolution of a particular research network, commissioned by a large research program. The project team developed a survey asking members of a newly formed research network to identify partners with whom they had worked with in the past year or so. The survey also asked if the collaboration (formal or informal) was a consequence of the newly formed research network. The information collected was processed with Excel and the UCINET software. With the analysis of the data and maps, the project team was able to develop a baseline for supporting the M&E strategy of the research program who commissioned the study. The characteristic of the network, including the characteristics of its members, their affiliations, disciplines, geographic distribution, areas of work and types of research conducted, to name a few, are likely to evolve with time as a consequence of the adjustments in the research collaboration and outcomes. The research program that commissioned the SNA study will monitor the evolution of the network by applying the same questionnaire and methods, on a periodical basis.
In both cases, SNA was used as a diagnostic and learning tool. The real benefit of SNA becomes apparent during and after the data collection and processing. SNA provides a learning opportunity as hundreds of data start taking shape and the findings are discussed with the network members. More about it at “Why use SNA?”, by TRAINMOR-KNOWMORE project.
For more information about SNA, I suggest the following resources:
This paper from the Asian Development Bank (ADB) focuses on the role of social networks and how analysis of these networks can benefit the groups and organisations within them.
This special edition of New Directions for Evaluation from the American Evaluation Association (AEA) edited by Maryann Durland and Kimberly Fredericks provides nine articles on social network analysis in program evaluation.
Our theme page on Network Evaluation lists many more resources. Also see this blog post by Simon Hearn from the EES conference which discusses the need to distinguish different types of networks and provides links to free SNA software.
Image source (top): Bartholomay, T., Chazdon, S., Marczak, M. S., & Walker, K. C. (2011)