Multilevel modelling workshop

Course summary

The multilevel modelling (a.k.a. hierarchical linear or mixed effects modelling) course is designed to extend regression models into the realm of multiple levels of analysis. Special attention is given to the translation of theoretical expectations into the statistical models, the interpretation and presentation of results and the general use and misuse of multilevel models in the social sciences (and how to specifically avoid the misuse part). While the course is predominantly designed to give you the knowledge of multilevel regression modelling and apply it to your research agenda, it does also arm you with the tools to run multilevel models in R. (If you prefer Stata, I do also have corresponding Stata code, but unfortunately little personal experience using it.)

Applications include models with continuous and limited dependent variables in hierarchical, longitudinal and cross-classified nesting situations and other advanced topics.

Format and activities

The course is a workshop, and therefore also include extensive discussions, tweaking and presentations of your own multilevel models. To get the most out of the course, come with your own multilevel modelling ideas and planned applications that we can turn into actual analyses during the week. Be ready to present the premise of your research already on the first day.

Prerequisite Knowledge

Advanced MA students with a clear application of multilevel models may also find the course suitable given appropriate prerequisite knowledge. The course is not  recommended to MA or more junior scholars who do not have a clear research application in mind. This course is designed for people with, at least, a solid foundation in regression models that reaches beyond knowing what to click to run a regression, how to copy the output into the paper, and knowing where to end the stars to point to in the write-up. Very basic knowledge of R is useful. At least know how to open a dataset and maybe run a linear regression, even if you are not yet ready to do all your data processing in R. If you have this much and are willing to work a bit more at it, we're in business. (And, again, while my examples will be using R, Stata equivalents will be provided as well.)

About the instructor

Levente Littvay (Levi) is Professor of Political Science at Central European University and a researcher of CEU's Democracy Institute. He researches survey and quantitative methodology, and the psychology of radicalism and populism. Award-winning researcher and teacher of graduate courses in research design, applied statistics, electoral politics, voting behavior, political psychology, and American politics. He was an academic convenor of ECPR's Methods Schools (2015-2021) and EUI's Fernand Braudel Fellow (2019-2020), Specialty Chief co-Editor for Methods and Measurement of Frontiers in Political Science and Associate Editor for Social Science of Twin Research and Human Genetics. Head of Team Survey in Team Populism and member of the ESS Round 10 Democracy and COVID modules questionnaire design teams. Recent books include Multilevel Structural Equation Modeling, and Contemporary US Populism in Comparative Perspective.

Reading list (indicative)


  • Luke, Douglas (2004) Multilevel Modeling. SAGE (Second edition is also OK.)
  • Singer, Judith, & John Willett (2003) Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. Oxford University Press


  • Brambor, Thomas, William Roberts Clark, & Matt Golder (2005), “Understanding Interaction Models: Improving Empirical Analyses", Political Analysis 13:1-20.
  • Andrew Bell, Malcolm Fairbrother, & Kelvyn Jones (2019), "Fixed and random effects models: making an informed choice", Quality & Quantity 53:1051-1074.
  • Enders, Craig K., & Davood Tofighi (2007), “Centering predictor variables in cross-sectional multilevel models: A new look at an old issue.", Psychological Methods 12(2):121-138.
  • Brian D. Johnson (2012) “Cross-Classified Multilevel Models: An Application to the Criminal Case Processing of Indicted Terrorists" Journal of Quantitative Criminology 28:163-189.
  • Alexander W. Schmidt-Catran, & Malcolm Fairbrother (2016) “The Random Effects in Multilevel Models: Getting Them Wrong and Getting Them Right" European Sociological Review 32(1):23-38.

More Fun MLM Readings:

  • Steenbergen, Marco, & Bradford Jones (2002), “Modeling Multilevel Data Structures.", American Journal Political Science 46(1):218-237
  • Stegmüller, Daniel (2013), “How Many Countries for Multilevel Modeling? A Comparison of Frequentist and Bayesian Approaches.", American Journal Political Science 57(3):748-761.
  • Martin Elff, Jan Paul Heisig, Merlin Schaeffer, & Susumu Shikano (2021) "Multilevel Analysis with Few Clusters: Improving Likelihood-Based Methods to Provide Unbiased Estimates and Accurate Inference" British Journal of Political Science 51:412-426.

Good Reference Texts (beyond the above textbooks):

  • Raudenbush, Stephen W. and Anthony S. Bryk. (2001) Hierarchical Linear Models: Applications and Data Analysis Methods (Second edition). Newbury Park, CA: Sage.
  • Hox, Joop (2010) Multilevel Analysis: Techniques and Applications, Second Edition. Routledge. (Newer editions are also OK.)