|
Date
|
Topic
|
Reading
|
Homework Due
|
|
Tuesday
April 1
|
Motivation
Overview of the use of social networks to model social structure important for understanding the spread of HIV.
|
Local Acts, Global Consequences: Networks and the Spread of HIV
|
|
|
Thursday
April 3
|
Introduction
Social science is the study of relationships and relationships can be
represented via social networks
|
Borgatti: Introduction to Social Networks
Radcliffe-Brown(1940): On Social Structure
|
|
|
Tuesday
April 8
|
Graph Theory and Notation
Nodes, ties (directed/undirected), degree, connectedness, cycles,
centrality, betweenness, etc
|
W&F Chapter 3
H&R Chapter 3
Borgatti (1994) (suggested)
|
|
|
Thursday
April 10
|
Data structures for representing graphs
Sociomatrix, edge list, R network data types, datasets, (bipartite,
affiliation)
|
W&F Chapter 4
H&R Chapter 4
|
Homework 1
Find paper on networks, read and summarize
|
|
Tuesday
April 15
|
Introduction to R
|
All about R notes
|
|
|
Thursday
April 17
|
Introduction to Networks in R
sna and network packages
reading and manipulating data, plotting, computing descriptive statistics
|
Sunbelt handout
|
Homework 2
Doing network descriptives by hand
|
|
Tuesday
April 22
|
Stochastic Models of Networks (estimation and
inference)
Model 1: Reney-Erdos: p(tie) is constant,
independent: joint distribution model, logistic model
Model 2: 2 different types of nodes with different probabilities of ties
Model 3: Vertex covariates model (logistic regression)
|
W&F Chapter 13.1-13.5
Snijders
(2003)
|
|
|
Thursday
April 24
|
Modeling Cohesive Subgroups
arbitrary mixing groups known a priori
likelihood inference
|
W&F Chapter 7
H&R Chapter 7
|
Homework 3
Use R (sna, network) to read data, descriptives, plots, centrality, etc.
|
|
Tuesday
April 29
|
Modeling Cohesive Subgroups continued
multiple groups unknown (latent class model)
Model 2: 2 different types of nodes with different probabilities of ties
Inference for models
|
Nowicki, K. & Snijders(2001). Estimation and prediction for stochastic
block models. Journal of the American Statistical Association, 96,
1077-1087
|
|
|
Thursday
May 1
|
Models for Fundamental Social Forces
1. Centrality (degree centrality, eigenvalue
centrality)
2. Sociality (undirected)
3. Prestige (directed)
4. Mutuality (directed)
|
W&F Chapter 5
H&R Chapter 6
|
Homework 4
Example Reney-Erdos, vertex attributes, mixing
|
|
Tuesday
May 6
&
Thursday
May 8
|
Modeling Cohesive Groups in Social Space
Network position (latent social space, probability of a tie proportional to
distance)
1-dimensional continuous observed
2-dimensional continuous observed
2-dimensional continuous unobserved
latent space cluster models
|
Hoff, Raftery, & Handcock (2001)
|
Homework 5
a) Latent class model
b) centrality model
|
|
Tuesday
May 13
&
Thursday
May 15
|
Introduction to general ERGM framework
general form
conditional independence models: Markov models, Hammersley-Clifford
1. Simulation of network via MCMC
2. Likelihood-based inference
3. Maximum likelihood and Bayesian inference
|
Hunter (2003)
|
Homework 6
Latent space models
|
|
Tuesday
May 20
|
Structure of triads: Triad Census
(Davis-Holland-Leinhardt)
transitivity
balance model (Heider)
Simmel model
|
W&F Chapter 14
|
|
|
Thursday
May 22
|
Mores sophisticated structural forms
cycles, triangles, gwsp, dsp,
esp, stars
|
Snijders et al (2006)
|
Homework 7
ERGM theory and MCMC, simulation of graphs
|
|
Tuesday
May 27
|
Goodness of fit of ERGMs
|
Hunter, Goodreau, & Handcock
(2006)
|
|
|
Thursday
May 29
|
Inference for partially observed networks
|
Handcock and Gile (2007), Gile and Handcock (2008)
|
|
|
Tuesday
June 3
&
Thursday
June 5
|
Sampling of networks (design)
ego-centered, link tracing
|
Gile and Handcock (2008); Frank (2004) Chapter 4
|
Homework 8
Triad census
Heider vs. Simmel
More sophisticated models
|
|
Extra
June
|
Network Dynamics
Summary
|
|
Homework 9
Goodness of fit sampling examples
|