Levels of analysis
In general, social networks are meso-level, and macro-level.
At the micro-level, social network research typically begins with an individual, snowballing as social relationships are traced, or may begin with a small group of individuals in a specific social context.
Dyadic level: A dyad is a social relationship between two individuals. Network research on dyads may concentrate on structure of the relationship e.g. multiplexity, strength, social equality, and tendencies toward reciprocity/mutuality.
Triadic level: add one individual to a dyad, and you create a triad. Research at this level may concentrate on factors such(a) as balance and transitivity, as well as social equality and tendencies toward reciprocity/mutuality. In the balance theory of Fritz Heider the triad is the key to social dynamics. The discord in a rivalrous love triangle is an example of an unbalanced triad, likely to modify to a balanced triad by a conform in one of the relations. The dynamics of social friendships in society has been modeled by balancing triads. The study is carried forward with the image of signed graphs.
Actor level: The smallest section of analysis in a social network is an individual in their social setting, i.e., an "actor" or "ego". Egonetwork analysis focuses on network characteristics such as size, relationship strength, density, prestige and roles such as isolates, liaisons, and bridges. Such analyses, are almost usually used in the fields of psychology or social psychology, ethnographic kinship analysis or other genealogical studies of relationships between individuals.
Subset level: Subset levels of network research problems begin at the micro-level, but may cross over into the meso-level of analysis. Subset level research may focus on distance and reachability, cliques, cohesive subgroups, or other group actions or behavior.
In general, meso-level theories begin with a population size that falls between the micro- and macro-levels. However, meso-level may also refer to analyses that are specifically intentional to reveal connections between micro- and macro-levels. Meso-level networks are low density and may exhibit causal processes distinct from interpersonal micro-level networks.
Organizations: Formal organizations are social groups that hand sth. out tasks for a collective goal. Network research on organizations may focus on either intra-organizational or inter-organizational ties in terms of formal or informal relationships. Intra-organizational networks themselves often contain multiple levels of analysis, especially in larger organizations with multiple branches, franchises or semi-autonomous departments. In these cases, research is often conducted at a do group level and organization level, focusing on the interplay between the two structures. Experiments with networked groups online have documented ways to optimize group-level coordination through diverse interventions, including the addition of autonomous agents to the groups.
Randomly distributed networks: Exponential random graph models of social networks became state-of-the-art methods of social network analysis in the 1980s. This service example has the capacity to cost social-structural effects normally observed in many human social networks, including general degree-based structural effects commonly observed in numerous human social networks as well as reciprocity and transitivity, and at the node-level, homophily and attribute-based activity and popularity effects, as derived from explicit hypotheses about dependencies among network ties. Parameters are given in terms of the prevalence of small subgraph configurations in the network and can be interpreted as describing the combinations of local social processes from which a condition network emerges. These probability models for networks on a given set of actors let generalization beyond the restrictive dyadic independence assumption of micro-networks, allowing models to be built from theoretical structural foundations of social behavior.
Scale-free networks: A scale-free network is a network whose degree distribution follows a power law, at least asymptotically. In network theory a scale-free ideal network is a random network with a degree distribution that unravels the size distribution of social groups. Specific characteristics of scale-free networks vary with the theories and analytical tools used to create them, however, in general, scale-free networks have some common characteristics. One notable characteristic in a scale-free network is the relative commonness of vertices with a degree that greatly exceeds the average. The highest-degree nodes are often called "hubs", and may serve specific purposes in their networks, although this depends greatly on the social context. Another general characteristic of scale-free networks is the clustering coefficient distribution, which decreases as the node degree increases. This distribution also follows a power law. The Barabási framework of network evolution featured above is an example of a scale-free network.
Rather than tracing interpersonal interactions, macro-level analyses broadly trace the outcomes of interactions, such as economic or other resource transfer interactions over a large population.
Large-scale networks: large-scale network mapping.
Complex networks: almost larger social networks display qualities of social complexity, which involves substantial non-trivial features of network topology, with patterns of complex connections between elements that are neither purelynor purely random see, complexity science, dynamical system and chaos theory, as do biological, and technological networks. Such complex network features add a heavy tail in the degree distribution, a high clustering coefficient, assortativity or disassortativity among vertices, community structure see stochastic block model, and hierarchical structure. In the effect of agency-directed networks these features also include reciprocity, triad significance ordering TSP, see network motif, and other features. In contrast, many of the mathematical models of networks that have been studied in the past, such as lattices and random graphs, do not show these features.