GESIS Training News

We wish you, your family and friends a happy and relaxing holiday season and a peaceful new year.

We very much appreciate your support, suggestions and feedback, and wish you all the best for 2019. We hope to see you again in Cologne or Mannheim next year at one of our events.

Your GESIS Training team

The GESIS Spring Seminar comprises three training courses for social scientists interested in advanced methods of quantitative data analysis. Lectures in each course are complemented by hands-on exercises giving participants the opportunity to apply these methods to data. For 2019, the GESIS Spring Seminar will focus on Bayesian Modelling in the Social Sciences.

Week 1 (March 11 - 15): Introduction to Bayesian Models for the Social Sciences

Prof. Dr. Susumu Shikano, University of Konstanz, Germany

Dr. Taehee Kim, University of Oldenburg, Germany

Week 2 (March 18 - 22): Bayesian Structural Equation Modelling

Prof. Dr. Rens van de Schoot, Utrecht University, Netherlands

Dr. Milica Miočević, Utrecht University, Netherlands

Week 3 (March 25 - 29): Bayesian Multilevel Modelling

Dr. Mark Andrews, Nottingham Trent University, United Kingdom

The courses can be booked either separately or as a block. There is no registration deadline, but places are limited and allocated on a first-come, first-served basis. To secure a place in the course(s) of your choice, we strongly recommend that you register now. We also recommend you to book your accommodation as soon as possible as rooms in Cologne can be scarce. Thanks to our cooperation with the Cologne Graduate School in Management, Economics and Social Sciences at the University of Cologne, doctoral students can obtain 3 ECTS credit points per one-week course.

Please register here (even if a course is fully booked)!

The GESIS Summer School will be held from August 01 to 23 at GESIS Cologne, Germany. 16 courses are scheduled, among them 4 short courses and 12 one-week courses. Here is an overview of this year's courses:

Week 0 (August 01 - 02): Short Courses

• Using Mplus for Latent-Variable Modeling: An Introduction
• Research Designs and Causal Inference
• Introduction to Data Analysis Using Stata
• Open Access to Research Data. Facing Funders’ Requirements on Making Research Data FAIR

Week 1 (August 05 - 09): Main Courses

• Introduction to Survey Design
• Introduction to Structural Equation Modeling: Confirmatory Factor Analysis with Mplus
• Questionnaire Design
• Mathematical Tools for Social Scientists: A Refresher Course with R

Week 2 (August 12 - 16): Main Courses

• Introductory Course to R with Applications from Data Analysis
• Mixed Mode and Mixed Device Survey
• Statistical Analysis of Incomplete Data
• Design and Implementation of Longitudinal Surveys

Week 3 (August 19 - 23): Main Courses

• Questionnaires for Cross Cultural Surveys
• Sampling, Weighting, and Estimation
• Pretesting
• Factorial Survey Design

For more information, please visit our website! Detailed course descriptions and the opening of the registration will be announced in January 2019 in our next newsletter.

Rens will conduct the course “Bayesian Structural Equation Modelling” within the frame of the GESIS Spring Seminar in March 2019 at GESIS Cologne.

How did you become interested in your subject?

I was always intrigued by scientists ignoring previous results when running new analyses. Why would you always start over again, I wondered, and ignore everything you already know? Then, I found out about Bayesian statistics which makes it possible to actually use background information and truly update your current knowledge when new data comes in.

What lessons can participants draw from your GESIS course?

Using previous knowledge can also be very dangerous, it makes your results ‘subjective’. But ignoring previous results is also highly subjective and dangerous. In this course, participants will learn when to worry when adding background knowledge to their models, and how to avoid the misuse of Bayesian statistics.

What do you think is the most exciting recent trend in Bayesian statistics?

Computers become faster almost every week and this opens up new opportunities to run complicated models. Moreover, all kinds of automated tools and checks are being developed so that applied researchers do not have to worry (too much) about what is going on under the hood of the Bayesian machinery. This way, more and more researchers are starting to use Bayesian statistics for analyzing their own data.

We thank Rens for his interesting insights.