Network Projects

• Project name: networks
• Research question (example): What is the relationship between gender and cross-program relations in high school?

• Expert contact: Javier Garcia-Bernardo

• Programming language: python (suggested) or R (allowed)

Canonical course conventions live in project_guidelines.md. That file is the source of truth for the four required workflow files (week1_explore.qmd, week2_operationalize_clean.qmd, week3_model.qmd, week4_storytelling.qmd), the data/model_data.rds -> data/model_results.rds pipeline, the raw-data policy, quality-check requirements, decision logs, and contribution tracking. Read it before starting and treat anything below as project-specific guidance on top of those conventions.

Tutorial framing

Network data are complex because observations are connected through ties, direction, weights, missing nodes, and dependence between relations rather than arriving as independent rows in a single analysis-ready table.

Students should learn three main things about these data:

1. How networks are represented through nodes, edges, edge lists, adjacency matrices, sparse matrices, GraphML, and choices about direction, weight, time, and isolates.
2. How to turn raw graph files into a clean network object while documenting what counts as a node, what counts as a tie, and which representation best matches the research question.
3. How network dependence affects standard statistical assumptions, and how network statistics, reference models, permutation tests, or clustering can support claims about homophily, polarization, centrality, or other network structures.

Peer-teaching checklist

Dimension	This project teaches
Data structure	Graphs with nodes, edges, node attributes, edge attributes, edge lists, adjacency matrices, and sparse matrix representations.
Storage system	File-based network repositories and downloaded graph files.
File formats	CSV edge lists, GraphML, and compressed repository downloads where relevant.
Encoding	Text CSV and XML-based GraphML.
Model	Assortativity or homophily statistic, permutation test, clustering, or a small network summary model.
Key aspects to explain	What counts as a node or tie, directed vs. undirected graphs, weighted vs. unweighted ties, isolates, sparse vs. dense matrices, network visualization, and why network dependence violates ordinary i.i.d. assumptions.

Resources

Data sources

• Network data:
  • Stanford Large Network Dataset Collection: https://snap.stanford.edu/data/
  • Network repository: https://networkrepository.com/networks.php
  • Netzschleuder: https://networks.sweked.de
  • Index of Complex Networks: https://icon.colorado.edu
• Potential dataset: https://networks.skewed.de/net/sp_high_school (interactive viz: https://javier.science/panel_network/)

Knowledge sources

• C/R/Python packages igraph,
• Guide for reference models: https://pubmed.ncbi.nlm.nih.gov/34216192/
• Observed network vs latent network: https://www.nature.com/articles/s41467-022-34267-9

Week-by-week

Week 1:

Begin with raw repository files and explain what the network is, who generated it, for what purpose, and the different storage formats.

• Explain the underlying network in substantive terms: what the nodes and ties represent, and whether the graph is directed or undirected, weighted or unweighted, static or temporal.
• What is GraphML? How does it relate to XML?
• Are adjacency matrices sparse or dense?
• Read about different layout algorithms.

Prepare for roundtable in week 2:

• What is a network and why is it a useful representation of data?
• What are the main ways to represent a network: edge lists, adjacency matrices, and XML or GraphML-like
• What are the advantages and disadvantages of adjacency matrices over edge lists? How do sparse matrices fix this and what are they?
• How do you visualize a network?

Week 2:

Operationalize the research question by turning raw graph files into a clean file with explicit decisions about direction, weights, and isolates.

• Think about the difference between network-wide questions like homophily or polarization and node-level questions like brokerage or node centrality.
• What are ways to measure the relationships between a node attribute (e.g. gender) and an edge attribute (cross-program relationships)?
• What is the “real” network?
• Commit to one primary network statistic for the project (e.g. assortativity by program/gender) and one permutation-style reference comparison. Do not try to cover multiple centrality measures, clustering, homophily, polarization, and temporal dynamics in the same project.

Prepare for roundtable in week 3:

• Be able to describe three analyses typically done on networks (e.g. assortativity, centrality, clustering) at a conceptual level, so the rest of the class understands the landscape — but your own project should report only the one statistic and one permutation comparison committed to above.
• Explain the selection vs influence debate in networks.

Week 3:

Use a network-appropriate statistic and an appropriate model, and check sensitivity to preprocessing choices.

• How does using different operationalizations of the network impact the result?
• Which parameters, specifically, answer our research question?

Prepare for roundtable in week 4:

• Explain how network dependence impacts standard i.i.d. assumptions
• Explain how the chosen network statistic or model works and what assumptions are hidden in the representation of ties.
• Explain how sensitive conclusions are to representation choices (e.g. type of network)

Week 4

Visualize and tell a story

• What is the context? What is the main result? Why is it important?
• Which visualizations support our research findings?
• What are the assumptions and limitations of your design?