Multilevel Modelling: Analyzing Cross-sectional and Longitudinal Survey Data

On this page: 
       Course description
       Target audience
       Course objectives
       Prerequisites
       Course content 
       Preparation for the course
       References
       Instructor’s Bio

Duration: 12 hours/3 days 

An hour-long lunch break from 13:00-14:00 with RECSM Winter School participants and members

Format

Hybrid (in-person and online)

Lecturer

Alexander Schmidt-Catran

Course Description

Many surveys spanning multiple countries--or many regions within a single country--are now being fielded multiple times over the course of years or even decades. Examples include the European Social Survey, International Social Survey Programme, EU Statistics on Income and Living Conditions, and (across states) the U.S. and German General Social Surveys. 
The range of topics that can be studied using data from these surveys is extremely broad: from health to religiosity to social and political attitudes and behaviours. This course will show students how to analyse these comparative longitudinal survey data (CLSD) using multilevel models that exploit any or all of three different kinds of variation: cross-sectional differences between countries, longitudinal change within countries over time, and variation across individuals.
We will begin by considering the hierarchical structure of CLSD and introducing the basic cross-sectional two-level model.  Next, we consider the longitudinal dimension and treat models for panel data. Then, we will see how CLSD can be understood as doubly hierarchical (or clustered), and therefore how we can analyse them with models partitioning between and within effects, making use of the lessons learned from panel data analysis.

Target Audience

This course is intended for postgraduate students, researchers and other social scientists who work with, or are interested in working with, comparative survey or panel data. Multilevel modelling is used across many disciplines, and the course will be relevant to participants studying topics such as education, health, social attitudes, political behaviour and social policy.

Course objectives

The course aims to provide students with a practical and conceptual introduction to the analysis of comparative longitudinal survey data using multilevel models. It will help participants understand the structure of these data, choose and interpret appropriate models, and apply the methods to their own research questions.

By the end of the course participants will be able to:

• Describe the hierarchical and longitudinal structure of comparative and longitudinal survey data.
• Explain the different questions addressed by fixed-effects and random-effects models.
• Distinguish between within-country and between-country relationships.
• Fit and interpret basic two- and three-level models, including models with random slopes.
• Use multilevel models to analyze change over time, including societal growth curves.
• Use graphical analysis to explore data and communicate results.
• Identify common analytical risks when modelling comparative longitudinal data.
• Relate course methods to a research question or project of their own.

Prerequisites

Students should have a basic understanding of regression models and some experience specifying and fitting them. Prior experience with multilevel models or panel-data analysis is useful but is not required. Students should be comfortable working with data and interpreting model output. No prior experience with R or Stata is required, although familiarity with statistical software will be helpful.

Course Content

Preparation for the course

Students can use R (main software) or Stata and need to bring their own laptop with the software already installed. For R, students should have the most recent version. All packages will be installed during the first session, where also the data will be provided. For Stata, the minimal required version is Stata 15, but newer versions are of course fine.

References

The course draws on the readings below. No advance reading is required; the papers are provided as recommended background and follow-up reading. Participants who would like to prepare in advance may begin with Bell and Jones (2015) and Fairbrother (2014).

• Bell, Andrew, and Kelvyn Jones. 2015. “Explaining Fixed Effects: Random Effects Modeling of Time-Series Cross-Sectional and Panel Data.” Political Science Research and Methods 3(1): 133–153. https://doi.org/10.1017/psrm.2014.7.
• Bell, Andrew, Malcolm Fairbrother, and Kelvyn Jones. 2019. “Fixed and Random Effects Models: Making an Informed Choice.” Quality & Quantity 53: 1051–1074. https://doi.org/10.1007/s11135-018-0802-x.
• Enders, Craig K., and Davood Tofighi. 2007. “Centering Predictor Variables in Cross-Sectional Multilevel Models: A New Look at an Old Issue.” Psychological Methods 12(2): 121–138.
• Fairbrother, Malcolm. 2014. “Two Multilevel Modeling Techniques for Analyzing Comparative Longitudinal Survey Datasets.” Political Science Research and Methods 2(1): 119–140. https://doi.org/10.1017/psrm.2013.24.
• Schmidt-Catran, Alexander W., and Malcolm Fairbrother. 2016. “The Random Effects in Multilevel Models: Getting Them Wrong and Getting Them Right.” European Sociological Review 32(1): 23–38. https://doi.org/10.1093/esr/jcv090.
• Shor, Boris, Joseph Bafumi, Luke Keele, and David Park. 2007. “A Bayesian Multilevel Modeling Approach to Time-Series Cross-Sectional Data.” Political Analysis 15: 165–181.

Instructor’s Bio

Alexander Schmidt-Catran is Professor of Sociology at Goethe University Frankfurt and a Visiting Professor at KU Leuven, Belgium. He teaches applied statistics, with a focus on multilevel models and the analysis of comparative and longitudinal data. He completed his studies and PhD in Cologne, Germany, and subsequently held a professorship in Mannheim before joining Goethe University Frankfurt. His methodological research on panel data and multilevel modelling has appeared in journals including Sociological Methodology, Sociological Methods & Research and the European Sociological Review. His substantive research focuses on attitudes towards the welfare state and immigration and has been published in journals including the American Sociological Review, the European Sociological Review and the Journal of Ethnic and Migration Studies.