Seminar Re-analyzing social network studies: What is the enemy of my enemy? (Winter 2015/2016)
|
13 January 2016 |
In this seminar we re-analyze social network data that has been used in published research. By varying the applied methodology we attempt to find out whether different methods of analysis would have lead to different results for the same research questions. The seminar, thus, touches the important issue of reproducibility in empirical social science. Participants will learn to apply simple and more sophisticated techniques to statistically analyze empirical social network data. We will re-analyze the hypotheses and findings published in Maoz et al. (2007) using data publically available from the Correlates of War Project (http://www.correlatesofwar.org/). In short, the paper addresses the question whether indirect relations, such as being an enemy of an enemy or being a friend of an enemy, have an influence on the occurrence of interstate conflict. In this seminar we are not interested in this research question per se but rather take it as a starting point to understand how the applied methodology affects the obtained findings. Prerequisites: Interest in quantitative methods for social network analysis and the ability to learn to work with statistical software, in particular R (http://www.r-project.org/). (Participants will get a short introduction and code examples.) It is not mandatory to have any previous knowledge about social network analysis, nor about social science, political science, or international relation research. Main reference: Zeev Maoz, Lesley G. Terris, Ranan D. Kuperman, and Ilan Talmud (2007). "What is the enemy of my enemy? Causes and consequences of imbalanced international relations, 1816-2001". Journal of Politics, 69(1):100-115. Note: There is a topically related lecture Network modeling. |
Schedule
| Weekly seminar (Viviana Amati & Jürgen Lerner) | Wed 10:00-11:30 in P 603 |
Tentative schedule of participants' presentations
| No. | Topic | Date | Presenter | Slides (pdf) |
|---|---|---|---|---|
| 1 | Repeat analysis of Maoz et al. | 25 November 2015 | - | - |
| 2 | Control for past alliances and conflicts | 2 December 2015 | Arno Fontaine | Pres. 2 |
| 3 | Control for covariates | 2 December 2015 | Dennis Soell | Pres. 3 |
| 4 | Analyze the probability of interaction | 9 December 2015 | Timon Behr | Pres. 4; Table1; Table2 |
| 5 | Analyze the conditional probability of conflict | 20 January 2016 | Qaiser Jamal | - |
| 6 | ERGMs for the conflict network | 16 December 2015 | Lena Pollak | Pres. 6 |
| 7 | ERGMs for the alliance network | 16 December 2015 | Dagmar Sorg | Pres. 7 |
| 8 | STERGMs for the conflict network | 20 January 2016 | - | - |
| 9 | STERGMs for the alliance network | 13 January 2016 | - | - |
| 10 | TERGMs with pseudolikelihood estimation | 20 January 2016 | - | - |
Material
Some documents are only locally accessible - see possibilities for remote access.Slides
- slides.pdf (last updated: 28 October 2015)
- Topic list: slides_topics.pdf (last updated: 28 October 2015)
- slides_regression.pdf (last updated: 11 November 2015)
Rscript
- Introduction to R (last updated: 27 October 2015)
- Statistical analysis with R (last updated: 03 November 2015)
- Specifying and estimating ERGMs with R (last updated: 11 November 2015)
Data
- International relations (Oneal and Russet, 2005) (file cmps2005.zip)
- Pre-processed data (file data_seminar.zip)
Software
Literature
Paper to be re-analyzed
- Zeev Maoz, Lesley G. Terris, Ranan D. Kuperman, and Ilan Talmud (2007). What is the enemy of my enemy? Causes and consequences of imbalanced international relations, 1816-2001. Journal of Politics, 69(1):100-115. (local copy)
International relations data
- John R. Oneal and Bruce Russett (2005). Rule of Three, Let It Be? When More Really Is Better. Conflict Management and Peace Science, 22:293-310. (local copy)
(Separable Temporal) Exponential Random Graph Models
- Hunter, D. R., Handcock, M. S., Butts, C. T., Goodreau, S. M., and Morris, M. (2008). ergm: A package to fit, simulate and diagnose exponential-family models for networks. Journal of statistical software, 24(3). (local copy)
- Krivitsky, P. N., and Handcock, M. S. (2014). A separable model for dynamic networks. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 76(1):29-46. (local copy)