Qualitative approaches: social network analysis

Social network analysis (SNA) is a method used to investigate social structures using network and graph theories. It characterises these structures by analysing nodes (individual actors, people or entities within the network) and the ties (relationships or interactions) that connect them. SNA operates on the principle that networked actors (i.e. nodes) and their actions are interdependent. The ties between actors serve as channels for the transfer or flow of resources (such as information, influence, money, etc.), with the network itself providing opportunities or constraints on the behaviour of the actors.

SNA can map and measure relationships and resource flows between actors, offering insights into bonding capital within a stakeholder network. It examines structural characteristics like the centrality or peripherality of specific actors and identifies emerging sub-networks that are only loosely connected to other parts of the network. SNA also highlights specific roles of actors within the network, such as boundary spanners between distinct sub-groups or between the network and external partners.

SNA helps assess the density, quality and robustness of communication structures between partners in formal and informal networks. The partners to be included in the SNA can be defined by the involved actors or the evaluators, depending on the purpose of the analysis. It is important to note that whereas SNA is an excellent method for visualising and calculating the structure of the networks and the roles of the actors, it only offers a snapshot of the network at a particular point in time.

Using SNA software, it is possible to create visual representations (network graphs) and conduct mathematical analyses of relationships between actors. These visualisations provide insights into the structure and dynamics of social networks.

SNA can explore networks in terms of actors and their relationships or interactions. For example, an SNA can be conducted on specific thematic aspects (e.g. the three CAP General Objectives or Specific Objectives), examining thematic network plots (e.g. for identifying key players) and on overlaps between them (e.g. for identifying key connectors), and discussing them in a focus group.

Other examples of social structures commonly visualised through SNA include social media networks, information circulation, peer-learner networks, business networks, knowledge networks and collaboration graphs. These networks are often visualised through sociograms, where nodes are represented as points and ties as lines. The size of each node indicates the relative importance of an actor in terms of the number of links, while the width of lines shows the intensity of interactions between organisations, and arrows indicate the direction of information flow. These visualisations provide a means of qualitatively assessing networks by varying the visual representation of their nodes and edges to reflect attributes of interest.

SNA provides a holistic image of social networks. Interpretations and judgements must be handled with care as the picture can be ephemeral, with communication structures shifting over time and circumstances. SNA diagrams can strongly stimulate group discussions by unveiling patterns that individual members of a social network might ignore.

In the context of the CAP Strategic Plan evaluation, SNA is valuable for assessing interventions where understanding the flow of information and relationships between different stakeholders is essential. This method is effective for monitoring and evaluating complex interventions involving multiple stakeholders and dynamic environments, such as Agricultural Knowledge and Innovation Systems (AKIS), LEADER and National CAP Networks (NN).

SNA can identify knowledge flows and stakeholders as knowledge providers, creators and users. It is important to consider that knowledge flows can occur at different levels of hierarchy. While institutions can act as knowledge providers in a top-down approach, SNA helps in understanding the horizontal knowledge flows between farmers.

SNA can collect evidence regarding indicators related to an intervention at two points in time, allowing for the calculation of changes in the average path length and the number of different types of actors involved. These changes must be related to the intervention under consideration through key informational interviews with knowledgeable, independent people (i.e. not involved in the intervention) who can verify or disregard causal claims.

For example, to assess the role of Operational Groups (OGs) in knowledge sharing, SNA can analyse OG plots to identify key players engaged in knowledge sharing within an OG, assess the structural characteristics of OG stakeholders (e.g. centrality or peripherality of stakeholders relevant to knowledge sharing within AKIS) and examine overlaps between them to identify key connectors. The results of these analyses can be discussed in a focus group.

To assess NN activities in terms of knowledge sharing, SNA can measure the involvement of relevant stakeholders and assess the effectiveness of network activities that support knowledge flows and exchanges. SNA can capture the capacity of the NN to facilitate the creation of sustainable networks (e.g. young farmers networks and school networks). For this purpose, it is essential to involve key actors early in the assessment.

In network evaluation, SNA can be used as a self-assessment exercise or by the evaluator to measure stakeholder involvement, monitor activities, assess the effectiveness of outputs, and plan further activities.

SNA can be applied in two different ways. First, by designing a questionnaire involving the application of SNA as a one-mode network to determine what should be known and measured. The second approach is the two-mode network, which can be applied using attribute data, eliminating the need for a questionnaire.

Data for SNA can be gathered through text analysis (e.g. meeting minutes), surveys and interviews. It is important to note that SNA requires relational data, meaning contacts, ties and connections, group attachments and meetings, which relate one stakeholder to another and cannot be reduced to individual stakeholders' properties.

Preconditions

The process of applying SNA in practice can include the following steps:

Step 1 – Define network boundaries

In Greece, SNA was used to reconstruct and understand the knowledge sources farmers personally assemble at various stages of their decision-making processes during the innovation process (e.g. awareness, assessment and implementation).

The study examined three types of innovations applied by farmers in rural areas characterised by different advisory landscapes, including the presence or absence and diversity of service providers (e.g. independent advisors, farmers’ social circles, producer cooperatives, research institutes and universities) and various advisory methods.

The cases regard the following innovations: (1) the implementation of an innovative integrated pest management practice involving mating disruption of insects using a network of micro-sprayers for peach cultivation (Imathia-Northern Greece); (2) the widespread cultivation of avocados (Chania-Crete); and (3) the introduction of stevia as an alternative crop to replace high-input and water-consuming traditional crops like tobacco and cotton with more profitable and environmentally friendly options (Karditsa-Central Greece).

SNA was integrated with concepts of social capital, ‘microAKIS’ (knowledge sources sought by farmers while innovating), and the Triggering Change Model, which explains farmers’ innovation decisions as major changes in their farming trajectory in response to trigger events.

Data collection involved a mix of methods, including surveys and in-depth interviews with 122 farmers. Farmers identified influential actors (suppliers of information, knowledge, and skills), the nature of interactions, and the frequency and direction of interactions.

The use of SNA helped to understand the diversity of actors from whom farmers seek advice during different innovation stages and the effectiveness of specific advisory methods in each case. SNA also identified: actors championing innovation or playing the role of network managers; the influence and frequency of different advice sources on farmers’ decision-making; the impact of specific advisory layouts on farmers’ construction of their microAKIS; triggering events preventing the adoption or leading to dropping innovations; and factors influencing decisions about adopting/non-adopting/dropping innovation.

See example:

Koutsouris, A., Zarokosta, E., (2021)
Farmers’ networks and the quest for reliable advice: innovating in Greece. Journal of Agricultural Education and Extension, Taylor & Francis (Routledge), 2021, p.1 - 27.