This workshop will aim to provide foundations in network theories and methods applied to studying relationships between variables or constructs. The examples will cover co-occurrence networks and other correlation-based network models often based on binary and/or continuous survey data. We will introduce the essential skills necessary to carry out the analysis in R, using mostly qgraph, igraph and other packages. The workshop will demonstrate several ways to estimate a network such as Gaussian Graphical Models, including the robustness checks and network comparisons. Furthermore, it will showcase adequate use and the interpretation of network analysis, such as centrality and community detection, to this specific type of data where nodes are not people (or other social entities), but rather concepts or variables. The descriptive analysis of estimated networks will cover cross-sectional data only.
The duration: 3.5 – 4 hours (with two breaks)
60% lecture; 40% practical
Students from different disciplinary backgrounds are welcome.
Basic knowledge of R is desirable, but not a requirement.