Social Network Analysis

What is Social Network Analysis?

Social Network Analysis (SNA) is a methodological approach used to analyze social structures through the use of networks and graph theory. It involves mapping and measuring relationships and flows between people, groups, organizations, computers, or other information/knowledge-processing entities.1

The Basic Idea

Imagine you're at a large conference, filled with a bunch of people. As you walk through the event and visit different rooms, you notice clusters of people invested in conversations, isolated individuals looking around, influential speakers drawing in crowds, and the event organizers making sure everything’s going according to plan. 

Every interaction, each connection, and each group dynamic is part of a larger and more complex web. This is, in a nutshell, what social network analysis (SNA) seeks to explore.2

SNA provides a way of visualizing relationships and their complex interplay by incorporating both quantitative and qualitative data. It aims to understand how different actors work together and share knowledge or resources across a network. From an organizational standpoint, SNA can help understand how information flows within a firm, identify key influencers in a social media community, or map the collaborations between scientists.

In essence, SNA involves creating a network map or graph where nodes represent actors (people, organizations, etc.) and the connecting edges or lines represent the relationships and interactions between them. Once these maps are drawn, analysts can find patterns such as clusters, central nodes, and bridges that link different parts of the network.2

How a Network Map Works

Going back to our conference scenario, let’s break down how a network map works. In a network map, each person at the conference is represented as a node (circle), and each interaction between them is represented as an edge (line). Thicker edges denote stronger relationships, such as long-term friends, while thinner edges denote weaker or newer connections.

  1. Nodes: Let’s say, Alice, Bob, Carol, David, Eve, and Tom are attendees (dark green). Emma and Liam are organizers (green). And Mike, John and Sarah are speakers (teal).
  2. Edges: The edges represent the interactions or relationships between nodes. Thicker lines represent stronger relationships. For example between Sarah and Mike, and Mike and Bob. Sarah and Mike know each other from work, and Bob and Mike are high school friends.

Connected nodes with different names and colors

By analyzing this map we can draw some conclusions. Sarah, Bob and Alice are known as high-degree nodes as they have the most interactions. In other words, they are key actors (or central figures). Organizers are well-connected and collaborate with one another. 

Tom is a bit of an isolated person who has only interacted with the organizers. Tom might benefit from being introduced to other attendees by his established connections, but we would need more information. Understanding these dynamics in this scenario could help plan more effective networking strategies for future conferences.

Theory, meet practice

TDL is an applied research consultancy. In our work, we leverage the insights of diverse fields—from psychology and economics to machine learning and behavioral data science—to sculpt targeted solutions to nuanced problems.

Our consulting services

Key Terms2

  1. Node: An individual entity within the network, such as a person, organization, or computer. For example, in a corporate network, each employee is a node.
  2. Edge: The connection or relationship between two nodes, representing interactions such as emails, meetings, or collaborations.
  3. Degree Centrality: A measure of the number of direct connections a node has. For instance, in a social network, a person with a high degree centrality might have many friends or followers.
  4. Betweenness Centrality: Known as gatekeepers as they facilitate or control the flow of information between other nodes.
  5. Closeness Centrality: A measure of how quickly a node can access other nodes in the network.
  6. Clustering Coefficient: A measure of the degree to which nodes in a network tend to cluster together.
  7. Community Detection: Identifying groups of nodes that are more densely connected internally than with the rest of the network.
  8. Whole Network Analysis: Type of SNA that observes connections between all actors in a network.3
  9. Ego-Centric Analysis: Type of SNA that focuses on understanding the network of a single actor or node.3


Social Network Analysis (SNA) can be traced back to the 1930s and was influenced by developments across various fields. Social networks were an intriguing topic, especially for sociologists and mathematicians. For example in 1934, Jacob Moreno investigated how an individual’s group relations impacted their actions and psychological development. His research led to the creation of sociograms — which would inspire social network maps as we know them today.

During the 1950s and 1960s, Dorwin Cartwright and Frank Harary evolved SNA by creating graph theory. This method introduces the concepts of positive & negative relationships, and group structure analysis from an individual perspective.

Meanwhile, researchers at Harvard, influenced by Kurt Lewin, introduced the concept of cliques, allowing network analysis to go beyond social structure description to actually identifying and understanding how different groups are structured.

The 1970s and 1980s were periods of significant development in network theory. Researchers like Mark Granovetter introduced the concept of the "strength of weak ties," highlighting the importance of weak social connections in spreading information. For example, you might have heard the phrase “it’s not what you know, but who you know.” Granovetter emphasized that weak ties usually come outside of your daily circle, allowing you to collect information that wouldn’t have been obtained otherwise.

Have you ever heard that there are six degrees of separation between yourself and anyone else around the world? Yes, there could be only six people between you and Taylor Swift. This hallmark research was conducted by Stanley Milgram, who demonstrated this theory using mathematics.

Once computers and the internet started being part of our daily lives, SNA became quite popular once again. Computers became nodes and interactions could happen anywhere, anytime, and between anyone. Additionally, tools such as UCINET, Gephi, and NodeXL make large-scale collection and analysis possible. SNA is now a tool used in fields ranging from epidemiology to marketing.4


Jacob L. Moreno: An Austrian-American psychiatrist, sociologist and philosopher known for creating visual representations of social relationships. His research helped explore the impact of group interactions on individual behavior and therapy.

Frank Harary: American mathematician who significantly contributed to the development of graph theory. His work introduced mathematical support to the study of social networks, allowing quantitative analysis of network structures and the relationships between nodes.

Kurt Lewin: A German-American pioneer of modern psychology who introduced the concept of cliques in network analysis. His research emphasized how social networks affect individual behavior and group performance.

Stanley Milgram: American psychologist who solved the "small-world problem" by realizing there were, on average, six degrees of separation between any two people.


As mentioned above, SNA is useful in multiple fields. For example, in public health, SNA has been crucial in identifying the spread patterns of diseases, allowing better interventions and more effective policies. During the COVID-19 pandemic, SNA was used to understand how the virus was spreading through different communities and countries. SNA was also leveraged to understand how news regarding the pandemic was spreading. Both of these analyses and strategies were useful in creating public policies, preventing the spread of misinformation and fear-mongering, as well as informing public health campaigns.5, 6

Another popular field is marketing. Some companies use SNA to identify content creators who can effectively promote their products or business. A content creators or influencers could have thousands or even millions of followers, however, your business would have to make sure those followers are the “right” followers or population for your business. So, by understanding content creators’ social networks, marketers can target specific users, maximizing their advertising results.7

In the corporate world, SNA is also used to understand how decision-making is being performed internally or externally, but also to analyze if some people are “over-networked”, meaning they have an excessive number of connections and interactions within the organization. This can lead them to be overwhelmed by the volume of responsibilities and at risk of burnout. In cases like this, SNA helps develop better teamwork, communication, and collaboration among employees and the company as a whole.2

SNA can be used to break down criminal networks by locating critical individuals and their relationships. Law enforcement organizations use social networks and communication trends to predict criminal activity and create intervention plans.


Ethical Concerns

One of the biggest ethical concerns in SNA is the potential for manipulating an individual's behavior. By identifying and influencing social networks, organizations or individuals can direct people’s decisions and actions. This could be especially problematic in political campaigns, where parties might manipulate people to their advantage. 

Another ethical issue is the lack of informed consent from individuals being analyzed. In theory, people accept terms and conditions when they create new social media accounts, but most people don’t read them. Therefore it might be legally safe to perform these studies, but not ethically as not all participants know they consented to their data being used. Promoting transparency and having regulations and ethical guidelines for SNAs ensures people are aware of how their data is being used.

Privacy Concerns

Also, when studies do collect this type of data, they collect personal and sensitive information about people’s behaviors. In the digital age, retaining anonymity and confidentiality can be quite challenging. It seems like there is a fine line between beneficial analysis and invasive surveillance. This can be quite significant in contexts such as health care, where patient confidentiality is crucial and breaking that trust leads to legal issues. Once again, transparency is paramount, but also implementing strong data encryption to ensure SNAs have effective anonymization methods that are not easily breached.

Potential Biases

Selection and Algorithmic Bias

Data used to analyze social networks are often not fully representative of the entire population, giving rise to selection bias. Unfortunately, this can lead to misguided conclusions and solutions. This type of bias is quite common in research as it’s difficult to observe an ENTIRE population (or even a subset that accurately reflects an entire population). This is why there are tools that help ensure data is as diverse and representative as possible. Also, being conscious and recognizing the bias that exists in research lowers ethical concerns and allows readers to understand the whole scope of your SNA. 

Taking this into consideration, algorithms can introduce biases into the analysis, particularly if data/participants were not selected properly. For example, although social media might sometimes seem like a gigantic cohort full of diverse contexts, it narrows down many other demographics. Having this bias can perpetuate inequalities and prejudices.

In conclusion, while SNA offers valuable insights and tools for understanding and influencing social dynamics, it also raises significant ethical, privacy, and practical concerns. Addressing these controversies requires a commitment to ethical principles, robust privacy protections, bias mitigation, and responsible use of SNA techniques.

Case Study

Using artificial intelligence, an analysis of 23,294 posts around the time of the 25th United Nations Climate Change Conference (COP25) in Madrid was performed to dissect the sentiments of tweets and social networks around Greta Thunberg and Bill Gates. This is what was known before performing the SNA:

  • Greta Thunberg, a young Swedish activist has become an important figure through school strikes and passionate speeches around climate change. Her messages are direct and urgent, often criticizing governments and corporations for not acting upon this matter. On the one hand, her activism has caused some people to have negative emotions towards her – expressed through tweets. On the other hand, her strong call to action has also pulled supporters and environmental activists who amplify her message through retweets.
  • Bill Gates, the tech mogul turned philanthropist uses a different approach. With the help of his resources and influence, Gates promoted long-term solutions and innovations to combat climate change. His tweets are usually received with a higher proportion of positive sentiments as he focuses on solutions and innovations. However, he seems to engage less with followers.

The SNA further highlighted these differences. Thunberg’s network is marked by intense interactions within tight-knit groups of activists and critics, creating a high degree of engagement but also significant opposition. Gate’s network, while larger in scope and less intense, is more cohesive in its support. Given that Gates is the synthesizer who advocates long-term systemic change, while Thunberg is the agitator who spurs rapid action; their differing techniques are reflected in the differences in network dynamics.

This case study illustrates how SNAs are relevant and crucial in understanding things such as differing approaches in digital activism, where both the messages and the way of delivering them play crucial roles in influencing public opinion and driving action.

Related TDL Content

Recurrent Neural Networks:
This article explains how recurrent neural networks (RNNs) process sequential data using internal memory, making them ideal for tasks such as natural language processing, speech recognition, and predictive text. It contrasts RNNs with feedforward neural networks, discusses the development and applications of RNNs, and addresses challenges and ethical considerations in their use.

The more the merrier? The irrationality behind group chats:

This article delves into the dynamics of group chats, exploring how they have transformed modern communication while also amplifying our inherent biases. It examines the formation and maintenance of these virtual communities, the cognitive limitations we face, and the biases that hinder our rational decision-making, offering practical steps to manage and optimize our digital social lives.


  1. Karl Blanchet, Philip James, How to do (or not to do) … a social network analysis in health systems research, Health Policy and Planning, Volume 27, Issue 5, August 2012, Pages 438–446,
  2. Visible Network Labs. (n.d.) Social Network Analysis 101. Retrieved on June 23rd, 2024 from:
  3. EEnet. (n.d.) What is a Social Network Analysis.  Retrieved on June 23rd, 2024 from:
  4. Zhang, M. (2010). Social Network Analysis: History, Concepts, and Research. In: Furht, B. (eds) Handbook of Social Network Technologies and Applications. Springer, New York, NY.
  5. Choudhury, T., Arunachalam, R., Khanna, A., Jasinska, E., Bolshev, V., Panchenko, V., & Leonowicz, Z. (2022). A Social Network Analysis Approach to COVID-19 Community Detection Techniques. International journal of environmental research and public health, 19(7), 3791.
  6. Pascual-Ferrá, P., Alperstein, N., & Barnett, D. J. (2022). Social Network Analysis of COVID-19 Public Discourse on Twitter: Implications for Risk Communication. Disaster medicine and public health preparedness, 16(2), 561–569.
  7. GRIN. (2023). How Network Analysis Helps in Finding the Right Influencers.  Retrieved on June 23rd, 2024 from:
  8. Ballestar, María Teresa & Llaguno, Marta & Sainz, Jorge. (2022). An artificial intelligence analysis of climate‐change influencers' marketing on Twitter. Psychology & Marketing. 39.

About the Author

Mariana Ontañón

Mariana Ontañón

Mariana holds a BSc in Pharmaceutical Biological Chemistry and a MSc in Women’s Health. She’s passionate about understanding human behavior in a hollistic way. Mariana combines her knowledge of health sciences with a keen interest in how societal factors influence individual behaviors. Her writing bridges the gap between intricate scientific information and everyday understanding, aiming to foster informed decisions.

Read Next

Notes illustration

Eager to learn about how behavioral science can help your organization?