Workshop sienaBayes

Sunbelt, Tuesday June 25, 2024

Location: Scott Russell SR 112

To prepare for these workshops:

• We probably will turn to do some work with actually running sienaBayes.
  You will be able to follow this better if you have a laptop with you, with RSiena installed and also multiSiena installed from the Siena and multiSiena downloads page.
• It will be good to have downloaded the RSiena manual.
• Take a look below at the program. There are some slides sets and handouts; perhaps you want to download some of them.

WS-T25: Introduction to the Analysis of Multilevel Network Dynamics using multiSiena.

First workshop, 9.00 AM - 12.00 PM

1. Pages 1-14 of Multilevel analysis of network dynamics (slides)
   ( and here are handouts) for the topics:
   • Multilevel concepts
   • Sample from Population of Networks
   • Four Approaches
   • Risks of having few groups
   • Use of Group characteristics
2. Principles of Bayesian inference
3. Pages 7-28 of Introduction to Multilevel Analysis of Network Dynamics using sienaBayes (slides)
   ( and here are handouts) for the topics:
   • Integrated hierarchical random effects models
   • Package multiSiena
   • The Bayesian model for multilevel SAOM
   • Data for sienaBayes
   • Very simple prior specification
   • Outline of estimation in sienaBayes
   • Operating sienaBayes
   • Results from sienaBayes
   • Continuing estimation and glueing together
   • Brief treatment of convergence assessment
4. At some moment during this treatment, we shall step to running sienaBayes for an elementary example:
   • Minimal intro to sienaBayes, perhaps to be replaced by
   • See github pages of Johan Koskinen
   • A further example is
     RscriptsienaBayes_A.r, an example script for the function sienaBayes in multiSiena, modeling network dynamics using a multi-group data set from Andrea Knecht's data, of which a description of one class is to be found here.
     
   • Perhaps also useful: Manual (i.e., help pages) for multiSiena.

WS-T30: Analyzing Multilevel Network Dynamics using multiSiena; further topics

Second workshop, 1.30 PM - 4.30 PM

1. Multilevel Analysis of Network Dynamics using sienaBayes (slides)
   ( and here are handouts) for the topics:
   • Model specification
   • Prior distributions
   • The function sienaBayes
     1. Bayesian MCMC procedure
     2. Parts of sienaBayes
     3. Initialization
     4. Numbers of iterations
     5. Convergence?
     6. Prolonging sienaBayes
     7. Further parameters
     8. Multiplication factor
   • Data exploration
2. At some moment we shall turn to one of the examples
   
   • RscriptsienaBayes_A.r, an example script for the function sienaBayes in multiSiena,
   • RscriptsienaBayes_B.r, an example script for the function sienaBayes in multiSiena,
   both using a multi-group data set from Andrea Knecht's data, of which a description of one class is to be found here.
   
3. Perhaps also useful: Manual (i.e., help pages) for multiSiena.

Materials

1. Websites :
   • The main Siena page
   • Siena and multiSiena downloads page
   • GitHub/stocnet
   • GitHub/stocnet/rsiena
   • Github pages of Johan Koskinen

2. Slide presentations :
   • Multilevel analysis of network dynamics (slides)
     Multilevel analysis of network dynamics (for printing)

   • Introduction to Multilevel Analysis of Network Dynamics using sienaBayes (slides)
     Introduction to Multilevel Analysis of Network Dynamics using sienaBayes (for printing)

   • Multilevel Analysis of Network Dynamics using sienaBayes (slides)
     Multilevel Analysis of Network Dynamics using sienaBayes (for printing)

   • An example of multilevel analysis of network dynamics using sienaBayes.
   • One-mode - two-mode co-evolution for studying social influence on behaviour , a presentation about the example in Koskinen & Snijders (2023) not highlighting the estimation by sienaBayes.
   • Goodness of Fit for Multilevel Dynamic Network Models (should be updated)
3. Scripts :
   • RscriptsienaBayes_A.r, an example script for the function sienaBayes in multiSiena, modeling network dynamics. using a multi-group data set from Andrea Knecht's data, of which a description of one class is to be found here.
     RscriptsienaBayes_B.r, a slightly more extensive example script, demonstrating the determination of the multiplication factor and the use of sienaTimeTest for the choice of randomly varying effects.
     RscriptsienaBayes_5.r, a more extensive example modeling dynamics of network and behavior and showing the use of endogenous group-level variables.
     
   • BayesPlots.r, a collection of plotting functions for sienaBayesFit objects, produced by sienaBayes.
   • The script used for Koskinen and Snijders (2023).
   • A script for changing network panel data to avoid impossible changes.
   • Another script for changing network panel data to avoid impossible changes (4 waves).
   • Some R objects produced by the scripts above; subdirectory Groups21 contains objects for RscriptsienaBayes_5.r.
4. Examples :
   • Isabel J. Raabe, Zsófia Boda and Christoph Stadtfeld (2019).
     The social pipeline: How friend influence and peer exposure widen the STEM gender gap.
     Sociology of Education, 92, 105-123.
     DOI: https://doi.org/10.1177/0038040718824095
     
   • Zsófia Boda, Bálint Néray, and Tom A.B. Snijders (2020).
     The Dynamics of Interethnic Friendships and Negative Ties in Secondary School: The Role of Peer-Perceived Ethnicity.
     Social Psychology Quarterly, 83, 342-362.
     DOI: http://dx.doi.org/10.1177/0190272520907594
     Multivariate and also multilevel.

5. Manuals :
   • RSiena manual; see Chapter 11.
   • Manual (i.e., help pages) for multiSiena.
6. Literature :
   • Johan Koskinen and Tom A.B. Snijders (2023).
     Multilevel longitudinal analysis of social networks.
     Journal of the Royal Statistical Society, Series A, 186, 376-400.
     DOI: https://doi.org/10.1093/jrsssa/qnac009
     This is an explanation of the statistical theory behind the sienaBayes function for multilevel longitudinal network analysis.
     It also presents an example of one-mode - two-mode network coevolution.
   • Tom A.B. Snijders (2016), The Multiple Flavours of Multilevel Issues for Networks.
     Chapter 2 in Emmanuel Lazega and Tom A.B. Snijders (eds.), Multilevel Network Analysis for the Social Sciences, Cham: Springer, 2016.
     ISBN 978-3-319-24518-8
     ISBN 978-3-319-24520-1 (eBook)
     DOI: 10.1007/978-3-319-24520-1