![](images/spacer.gif)
![](images/spacer.gif)
![](images/spacer.gif)
Subscribe & download
![](images/spacer.gif)
Filter events
|
![](images/spacer.gif)
3:30pm to 4:30pm |
|
Statistics Colloquium on Social Networks
(Academic)
Please join us for a special Department of Statistics colloquium on Social Networks.
"Generalized Hierarchical Models for Relational Data"
Andrew C. Thomas
Harvard University Dept. of Statistics
A great increase in relational data, observations of interactions between individuals, has made network analysis methods of great interest to scientists of all types. As most commonly used methods for analysis of relations require the data to be binary in nature, it is essential to use methods that account for the magnitude of an observation. To that end, we introduce hierarchical linear modelling approaches to relations with numerical outcomes in order to obtain meaningful inferences on relational ties in a network as well as of the individuals comprising it. We demonstrate our methods on the number of press releases sent jointly by United States senators across several years.
This work is in collaboration with Prof. Joseph Blitzstein of the Harvard Statistics Department and Justin Grimmer of the Harvard Government Department.
|
|