A Look Back – GESIS Fall Seminar in Computational Social Science 2024
From 30 August to 27 September 2024, the GESIS Fall Seminar in Computational Social Science took place in Mannheim and online. For the first time, it offered not only live courses but also two courses in our brand new blended learning format. Overall, participants could choose from nine introductory and advanced courses on computational social science methods and techniques.
Designed to accommodate diverse schedules, our “Introduction to Computational Social Science” courses with both R – taught by Johannes R. Gruber (Vrije Universiteit Amsterdam) – and Python – taught by John McLevey (University of Waterloo) – gave an overview of different methods, techniques, and tools for the collection and analysis of digital behavioral data using a blended learning format that combined guided self-learning phases with live online classroom sessions.
Those who wanted to dig deeper into the data collection process could delve into web scraping methods and how to work with APIs in our bilingual course on “Web Data Collection with R and/or Python,” facilitated by Iulia Cioroianu (University of Bath) and teaching assistants David Schweizer and Georg Ahnert (both University of Mannheim).
Focusing on the analysis of textual data, Marieke van Hoof (University of Amsterdam) and Rupert Kiddle (Vrije Universiteit Amsterdam) gave an “Introduction to Machine Learning for Text Analysis in Python.” In “Introduction to Social Network Analysis with R,” Philip Leifeld (University of Manchester) showed participants how to manage, visualize, and analyze social network data.
Michael Mäs and Fabio Sartori (both Karlsruhe Institute of Technology) provided a primer and overview of the principles of complexity science in “Agent-based Computational Modeling.” More advanced participants had the chance to learn about word embeddings and state-of-the-art large language models like GPT from Hauke Licht and Lisa Maria Wierer-Lechner (both University of Innsbruck) in “From Embeddings to LLMs: Advanced Text Analysis with Python” or explore statistical network models like ERGMs and SAOMs in “Advanced Social Network Analysis with R” with Michał Bojanowski (Kozminski University and Universidad Autònoma de Barcelona).
In their live online course, Andreu Casas (Royal Holloway University of London) and Felicia Loecherbach (University of Amsterdam) shared their expertise on how to use computer vision and machine learning for “Automated Image and Video Data Analysis.”
Held on site in Mannheim as well as streamed live to remote participants, two expert talks by Marc Ratkovic and Ines Rehbein (both University of Mannheim) on “Integrating LLMs into Quantitative Social Science” and “Potentials and Pitfalls of Transformer-based Text Classification for the Social Sciences,” respectively, provided food for thought on the future of computational text analysis and a fruitful basis for discussion.
Participants in Mannheim further enjoyed the opportunities for networking at the weekly get-together at GESIS and the dinner at Rheinterrassen.
Not only the social environment but also all courses were rated very highly by participants: More than 9 out of 10 participants stated they were (very) satisfied with their course, and a whopping 94% said they would recommend GESIS Training events. Participants particularly emphasized the vast knowledge of the lecturers, their readiness to give individual feedback and support, and the hands-on nature of the courses. They also greatly appreciated that the courses “[brought] together theory and method,” provided the opportunity for “thinking about our own research design,” and gave “very useful tips and information.” To quote one participant in “Introduction to Machine Learning for Text Analysis in Python”: “[The] teachers were very proficient, kind, patient, helpful: score 11 out of 10!” A big THANK YOU to our fantastic lecturers for their commitment and to our participants for placing their trust in us!
Save the date: The next GESIS Fall Seminar in Computational Social Science will take place from 1 to 26 September 2025 . We will announce the full program in Spring 2025 (www.gesis.org/fallseminar).
Stay tuned and subscribe to our newsletter at https://www.gesis.org/en/gesis-training/about-us/newsletter!
