Penn State boasts one of the strongest political methodology faculties in the country, developing new and innovative approaches to data collection, data analysis and statistical inference. Faculty specialties include methods of causal inference (including experimental and quasi-experimental design), machine learning and textual analysis, semi-and non-parametric models, survival and time series analysis, multivariate statistics, event data analysis, formal modeling, game theory, and survey research.
Our Social Data Analytics program emphasizes the need for emerging methods in data science to be informed by theoretical and substantive expertise on politics, society, and human behavior.
Faculty Research
Improving statistical analysis
- How should we specify and interpret empirical models to test conditional claims about the world? [Matt Golder]
- How should analysts incorporate uncertainty over the dynamic properties of individual time series when testing hypotheses about the relationship between them? [Suzanna Linn]
- Text analysis & machine learning
- Can we develop text analytic techniques to extract richer measures of political opinion from open-ended survey questions? [Burt Monroe]
- How can one build a fully-unsupervised sentiment dictionary that is adaptive to specialized vocabulary in non-standard corpuses? [Chris Zorn]
- How do we blend the tools and standards of social science and data science to benefit from each other? [Burt Monroe]
- How can we use social media data to augment public opinion polls and other probability surveys? [Eric Plutzer|
Network analysis
- What can we learn about political candidates based on the networks of donors who finance their campaigns? [Bruce Desmarais]
- How can we best identify the diffusion of political beliefs and behaviors through networks? [Bruce Desmarais]
- How can we model the dynamics of complex political networks (e.g., collaboration among legislators, communication among administrative officials, conflict between nations) [Bruce Desmarais]
- Formal models
- How can Spatial Cure Models improve our understanding of the spread of regime change? [Bumba Mukherjee]