Social network analysis digs into the connections between people, groups, and organizations. It looks at how these relationships form patterns and influence behavior, using concepts from network and graph theories to make sense of complex social structures.

Key elements include actors (nodes), ties (edges), and attributes. Analysts examine network structure, actor roles, and how networks change over time. This helps understand how connections impact outcomes like performance, innovation, and information spread.

Fundamental Concepts and Principles

Fundamentals of network analysis

• Social network analysis (SNA) studies social structures and relationships using network and graph theories
  • Focuses on patterns and implications of connections between individuals, groups, or entities (friends, colleagues, organizations)
• Key elements of social networks
  • Actors or nodes represent individuals, organizations, or entities within the network (people, companies, websites)
  • Ties or edges represent relationships, interactions, or connections between actors (friendships, collaborations, hyperlinks)
  • Attributes are characteristics or properties associated with actors or ties (age, gender, tie strength)
• Social network data typically represented using adjacency matrices or edge lists
• SNA aims to understand and analyze
  • Network structure and topology (centrality, density, clusters)
  • Roles and positions of actors within the network (brokers, influencers, isolates)
  • Dynamics and evolution of networks over time (growth, diffusion, resilience)
  • Impact of network properties on individual and collective outcomes (performance, innovation, contagion)

Network Measures and Analysis

Centrality measures for influence

• Centrality measures quantify the importance or influence of nodes within a network
• Degree centrality
  • Measures the number of direct connections a node has
  • Nodes with high degree centrality are considered hubs or connectors (celebrities, airports)
• Betweenness centrality
  • Measures the extent to which a node lies on the shortest paths between other nodes
  • Nodes with high betweenness centrality are important for information flow and control (bridges, gatekeepers)
• Closeness centrality
  • Measures the average shortest path distance from a node to all other nodes
  • Nodes with high closeness centrality can quickly reach or influence others (central locations, rapid responders)
• Eigenvector centrality
  • Measures the influence of a node based on the importance of its connections
  • Nodes connected to other influential nodes have higher eigenvector centrality (PageRank, prestige)

Network structure and dynamics

• Graph theory provides mathematical foundations for representing and analyzing networks
  • Graphs consist of vertices (nodes) and edges (ties) connecting them
  • Network properties derived from graph metrics (density, diameter, average path length)
• Network visualization techniques explore and communicate network structures and patterns
  • Force-directed layouts position nodes based on their connections, revealing clusters and central actors
  • Node and edge attributes mapped to visual properties (size, color, shape)
  • Interactive features allow exploring subsets or ego networks around specific nodes
• Community detection algorithms identify densely connected subgroups or clusters within a network
  1. Modularity optimization methods maximize within-group connections and minimize between-group connections
  2. Hierarchical clustering reveals nested community structures at different resolution levels
• Network dynamics studied by analyzing changes in network structure over time
  • Temporal network analysis tracks formation, dissolution, or evolution of ties
  • Stochastic actor-oriented models (SAOMs) simulate network dynamics based on actor attributes and local network configurations

Properties of social media networks

• Network density measures proportion of actual connections relative to maximum possible connections
  • High-density networks have many connections and facilitate information diffusion (small groups, team projects)
  • Low-density networks have fewer connections and may have structural holes or brokerage opportunities (large organizations, international trade)
• Clustering coefficient quantifies tendency of nodes to form tightly connected groups or triangles
  • High clustering indicates presence of cohesive subgroups or communities (friend groups, research collaborations)
  • Social media networks often exhibit high clustering due to shared interests or affiliations
• Homophily refers to tendency of individuals to associate with similar others
  • Homophily based on attributes (demographics, attitudes, behaviors)
  • Social media networks may exhibit homophily, leading to echo chambers or reinforcement of existing beliefs (political discussions, brand communities)
• Network properties have implications for information diffusion, social influence, and collective behavior
  • Dense and clustered networks can accelerate spread of information or behaviors (viral marketing, innovation adoption)
  • Homophilous networks may limit exposure to diverse perspectives and reinforce polarization (filter bubbles, group polarization)
  • Influential nodes identified through centrality measures targeted for marketing, interventions, or opinion leadership (influencer marketing, public health campaigns)