44756 - Social Networks Analysis

Learning outcomes

People credited with this course are able to explore a process in the social, economical and technical fields and to model a “real life” problem by a network approach. That implies: i) to know the theoretical SNA - Social Network Analysis framework (concepts about network components and structure, network statistic, network representational power) and ii) to acquire some practical skills in node-ties identification; data identification and collection; dataset building and data handling; measurements choice, carrying out and interpretation; graphical representation by the most common statistical net-packages.

Course contents

• Social Network Analysis in the fields of social and economic sciences: theoretical background. Differences with the traditional quantitative approaches in social analysis.

• Network definitions. Network data: nodes and ties representations. One-mode and two-mode networks. Ego-centered networks. Directed, signed and valued graphs. Adiacency and incidency matrices. Hypergraphs. Multigraphs. Matrices algebraic operations and their meanings. Blockmodel and image matrices. Nodes attributes.

• Descriptive network statistic. Nodal degree: in e out degree, density, reachability, connectivity, walks and paths, distance, geodetic distance, diameter, maximum flow, reciprocity, transitivity and simmelian ties, clustering coefficent, External-Internal index.

• Concept of centrality. Relations between centrality and power. Centrality measures: degree centrality (Freeman and Bonacich measures), closeness centrality, betweenness centrality. Eigenvector. Hubbel and Katz influence measure.

• Groups and subgroups. Cliques, N-cliques, N-clans, K-cores, F-groups. Principal components analysis. Cutpoints and bridges.

• Role and position concepts. Similarity/Dissimilarity. Structural, automorphic and regular equivalence.

• Egonet: definition and measurement. Structural holes: Burt theory. Brokerage: Fernandez and Gould measures.

• Small words ad scale-free networks: properties, algoritms. Concept of resilience.
  

• Two-mode networks: “actor/event” relations. Data modelling. Affiliation. Some two-mode statistics.

• Network graphical representations: layouts. Cognitive bias of interpretation.

• Design of a network analysis: problem modelling, data identification and collection, dataset building and data handling, measurements choice, validity and interpretation. Privacy implications.

Readings/Bibliography

Teacher lecture notes. Network dataset for training use.

General bibliography
Koch R. and Lockwood G. (2010), Superconnect. The Power of Networks and the Strenght of Weak Links, London, Little Brown.
Watts D. (2004), Small Worlds. The dynamicof Networks between Order and Randomness, Princeton, Princeton University Press.
Carrington P., Scott J. and Wasserman S., (2005), Models and Methods in Social Network Analysis, Cambridge (MA), Cambridge University Press.
Wasserman S. and Faust K. (1996), Social Network Analysis. Method and Applications, Cambridge (MA), Cambridge University Press.
Barabasi A-L. (2003), Linked: How Everything Is Connected to Everything Else and What It Means, Cambridge (MA), A plume book.

Teaching methods

Frontal teaching alternated with classroom exercices.

Assessment methods

Individual or small group project work, negotiated with the teacher

Teaching tools

Assisted case studies applying several statistical net-packages (Ucinet 6.90, Pajek 1.02, Netdraw 2.12, Mage 6.02, Visone 2.2). Group design of real applications.

Office hours

See the website of Marco Ruffino