GESIS Training Courses
user_jsdisabled
Search

Scientific Coordination

Sebastian E. Wenz
Tel: +49 221 47694-159

Administrative Coordination

Loretta Langendörfer M.A.
Tel: +49 221 47694-143

Please wait...

Advanced Modeling of Categorical Dependent Variables

About
Location:
Hybrid (Online via Zoom / Cologne; Unter Sachsenhausen 6-8)
 
General Topics:
Data Analysis
Course Level:
Advanced
Format:
Spring Seminar
Software used:
R
Duration:
5 days
Language:
English
Fees:
Students: 550 €
Academics: 825 €
Commercial: 1650 €
 
Keywords
 
 
 
 
 
Additional links
Lecturer(s): Maria Kateri, Irini Moustaki

About the lecturer - Maria Kateri

About the lecturer - Irini Moustaki

Course description

This course offers a comprehensive and application-oriented introduction to the statistical modeling of categorical variables, tailored to the needs of researchers in the social and behavioral sciences. Categorical outcomes, such as survey responses, ratings, or classifications, are widespread in these fields but require models that go beyond standard linear regression. The course begins with the fundamentals of contingency tables and models for contingency tables, covering log-linear models, and models for ordinal classification variables, with special attention to model selection, fit assessment, and interpretation. We then move on to generalized linear models, focusing on regression modeling with categorical dependent variables, including binary logistic regression (logit, probit, and complementary log-log models), multinomial regression, and ordinal regression, along with strategies for handling sparse data and separation issues. Advanced sessions introduce multilevel models for categorical outcomes and approaches for analyzing longitudinal and clustered data with mixed-effects logistic models. The final day is devoted to models with latent variables such as factor analysis and latent class analysis for categorical observed variables. Each day includes hands-on lab sessions, where participants will apply the discussed methods to real datasets using specific R packages. This course is designed for researchers and advanced students with basic statistical training who wish to deepen their methodological toolkit for analyzing categorical data in empirical research.
 
The complete syllabus for this course will shortly be available for download here.
 
Organizational structure of the course
Each day of the course is structured into two main parts: a theoretical session lasting approximately 3-4 hours, followed by a practical R lab of 2-3 hours. During the theoretical sessions, core concepts, methodologies, and models will be presented. In the R labs, students will have the opportunity to implement these methods themselves, gaining hands-on experience and reinforcing their understanding through practical application.


Target group

You will find the course useful if:
  • You are a researcher or advanced graduate student in the behavioral or social sciences working with categorical or ordinal data from surveys, experiments, or observational studies.
  • You are applying binary or ordinal regression models and are interested in advanced topics such as multilevel models or models with latent variables.
  • You are interested in implementing statistical models using R.


    • Learning objectives

    By the end of the course you will:
  • Be able to critically evaluate and select appropriate statistical methods for analyzing categorical and ordinal data, based on the context and research objectives.
  • Learn to implement and interpret generalized linear models such as logistic, and ordinal regression.
  • Learn to conduct and interpret contingency table analyses, address challenges like sparse data and model diagnostics, and apply multilevel and longitudinal modeling techniques.
  • Develop skills in factor models and latent class models for categorical data.
  • Gain practical experience with relevant R tools for data analysis and model implementation.
  • Strengthen your ability to interpret and effectively communicate the results of a statistical analyses of categorical data.
  • Be prepared to apply advanced statistical techniques for categorical data to real-world data in your own research or professional work.


    • Prerequisites

  • Basic knowledge of statistical concepts (e.g., probability, basic distributions for random variables, hypothesis testing, regression analysis) and basic methods of statistical analysis for categorical data.
  • Some familiarity with generalized linear models (GLMs), especially logistic regression.
  • Basics on R programming for data analysis and model implementation.
  • Some exposure to multilevel or longitudinal data structures is helpful but not mandatory.
    •  
    Software and hardware requirements
    The course will involve hands-on exercises using R. As a participant, you should bring your own laptop for use in the course. Please install recent versions of R (≥4.3.0) and RStudio before the course. The latest versions of R and RStudio are available for free at https://cran.r-project.org/ and https://www.rstudio.com/.
     
    Please make sure that you are able to download files from the internet - free Wifi is provided by GESIS - and that you have the rights to install R packages on your laptop during the course. We will provide a list of the packages to be installed shortly before the workshop.


    Schedule

    Recommended readings