Introduction to Social Network Analysis

Key topics covered

The course distinguishes the production of relational data from their analysis. The opening sessions address how such data are generated, across survey, archival, administrative, digital, and mixed-methods sources. The remaining sessions are devoted to analysis and inference.

Two designs are treated. In the complete network design, every tie among a bounded set of actors is recorded and the network is the object of study. In the personal network design, each respondent reports their own ties, and the network is an attribute of the individual. Days 2 and 3 adopt the first design and Day 4 the second, on the same data.

Days 2 to 4 are descriptive: position, structure, and groups in complete networks; size, density, and composition in personal networks. Day 5 is inferential: hypothesis testing, the dependence among relational observations, and its consequences for estimation.

Social networks appear on both sides of explanation, as explanans and as explanandum. On Day 4, position and composition enter a regression as explanatory variables. On Day 5, the tie is the dependent variable, and four techniques are introduced: the quadratic assignment procedure, exponential random graph models, stochastic actor-oriented models, and relational event models.

Each session comprises two taught blocks and a laboratory in R. Annotated code and prepared data are supplied for every exercise. Office hours are available throughout the course week.

Day 1: Concepts, boundaries, and relational thinking

Nodes, ties, direction, weights, and one-mode and two-mode networks; graphs and matrices as two representations of the same object. Boundary specification, sampling, and the consequences of missing ties. Visualisation is treated as a stage of the analysis: what a layout algorithm conveys and what it leaves arbitrary. Laboratory: read a network into R, inspect it, and produce an annotated plot with attributes mapped to colour and size.

Day 2: Where relational data come from, and first measures

Survey instruments: name generators, name interpreters, and position and resource generators. Archives and documents. Administrative records and digital traces. Networks in mixed-methods research, with ties elicited through interviews and fieldwork. What each source supports, and the ethical problems specific to relational data. Degree, closeness, and betweenness. Laboratory: build a network object, navigate it, and compute the first centralities.

Day 3: Position, structure, and groups

Eigenvector centrality, where importance depends on the importance of one’s contacts. Brokerage as a distinct form of advantage, and what each measure captures. Density, components, path length, clustering, centralisation, and community detection. Laboratory: compare centralities on one network, describe it as a whole, and justify a choice of measure.

Day 4: Personal networks

The individual as the unit of analysis, and the limits of what egocentric designs observe. Size, density, and composition, including homophily and diversity; constraint and effective size. Ego-level measures as covariates in a regression, and the separation of selection from influence. Laboratory: navigate personal network data, assemble a file with one row per ego, and fit the regression.

Day 5: Modelling relational data

Observations in a network depend on one another; the consequences for standard errors, and permutation-based inference. The tie as dependent variable. The quadratic assignment procedure, exponential random graph models, stochastic actor-oriented models, and relational event models, each comparing the observed network with those that could have been observed. Laboratory: fit one of the models introduced and state what it identifies.

Post-course assignment

You can choose between two formats. Each requires roughly 25 hours of work and runs to between 3,000 and 5,000 words, excluding code, tables, and figures.

• An empirical analysis. An original analysis of relational data, your own or one of the datasets supplied during the course. State a research question and locate it in the literature, describe how the data were collected and what that implies for the claims available, present a visualisation and descriptive measures, and estimate at least one test or model. Submit the R code with the text.
• A research design. If you have yet to collect relational data: a network study you could realistically carry out, specifying the question and its place in the literature, the source or instrument, the boundary, the measures that source supports, the analysis, and the threats to inference you anticipate.

Both formats require an explicit account of data collection and its consequences for what can be claimed.

Prerequisite knowledge

Network analysis itself is taught from first principles. You should be comfortable with basic quantitative methods: distributions, means, measures of association, and the logic of linear and logistic regression. The final session uses network measures as variables in a regression, and asks you to read and interpret the model.

You should also have a basic working knowledge of R and RStudio: opening a script, running it, installing a package, recognising a data frame. Annotated code is supplied for every exercise. If you have yet to use R, R for Data Science and Data Science in a Box are good places to start.

You will need R and RStudio, both free and open source, installed on your machine before the start of the course.

Before the course

Read these four texts before the first session. Together, they set out the logic of social network analysis:

• Borgatti, Everett, Johnson and Agneessens (2022), chapters 1 and 2: what a network is, and what network analysis does.
• Borgatti, Mehra, Brass and Labianca (2009): how far the approach has travelled across the social sciences.
• Ward, Stovel and Sacks (2011): what network analysis has done in political science.
• Emirbayer (1997) on the relational turn.

If you haven't already, you will also install the software. If R is new to you, look at R for Data Science and Data Science in a Box. Allow at least four hours.

Readings

• Borgatti, S.P., Everett, M.G., Johnson, J.C. and Agneessens, F. (2022) Analyzing Social Networks Using R. Los Angeles: SAGE.
• Borgatti, S.P., Mehra, A., Brass, D.J. and Labianca, G. (2009) ‘Network Analysis in the Social Sciences’, Science, 323(5916), pp. 892–895. doi: 10.1126/science.1165821.
• Emirbayer, M. (1997) ‘Manifesto for a Relational Sociology’, American Journal of Sociology, 103(2), pp. 281–317. doi: 10.1086/231209.
• Ward, M.D., Stovel, K. and Sacks, A. (2011) ‘Network Analysis and Political Science’, Annual Review of Political Science, 14. doi: 10.1146/annurev.polisci.12.040907.115949.

Learning commitment

The course includes three hours of live teaching each day for five days, alongside readings and other learning activities.

Once you have registered and paid, you will receive access to our Learning Management System (LMS) approximately two weeks before the course start date. Course materials, including pre-course readings, will be available here. Allow approximately 20 hours per week before the course to review the materials and complete readings and any pre-course tasks.

During the course week, you should allow approximately two to three hours per day for preparation and assignments, in addition to the live teaching sessions.

The course offers the opportunity to earn three ECTS credits. To earn a fourth credit, you will need to complete a post-course assignment involving approximately 25 hours of work.

Disclaimer

This course description may be subject to subsequent adaptations (e.g. taking into account new developments in the field, participant demands, group size, etc.). Registered participants will be informed at the time of change.

By registering for this course, you confirm that you possess the knowledge required to follow it. The instructor will not teach prerequisite items. If in doubt, please contact us before registering.