Workshop by Dr João Veríssimo: Analyzing individual variability
MGK in collaboration with Project INF and GRADE Center Language will host a three-day workshop on analysing individual variability in linguistic and psycholinguistic experiments, led by Dr. João Veríssimo (University of Lisbon). The workshop will take place on October 5-7, 2026, from 10:00 to 16:00 in room IG. 251 and via Zoom. To register, please contact grade-language@em.uni-frankfurt.de by September 15, 2026. The Zoom link will be shared with registered participants.
Date: October 5-7, 2026
Time: 10:00-16:00
Location: IG. 251 and via Zoom (all three days)
Presenter: Dr. João Veríssimo, University of Lisbon
Registration: grade-language@em.uni-frankfurt.de
Registration deadline: September 15, 2026
Analysing Individual Variability Using Bayesian Hierarchical Models
This hands-on workshop provides a practical tutorial on analysing individual variability in linguistic and psycholinguistic experiments using Bayesian hierarchical models. By the end of the workshop, participants will have developed a comprehensive understanding of how different sources of variability can be modelled and analysed and will be able to apply these methods to their own datasets.
Data from linguistic and psycholinguistic studies often display substantial individual-level variability: participants may differ widely in both their overall performance and their responses to experimental manipulations. Hierarchical models provide a principled way to estimate between-participant variability while distinguishing it from within-participant variation. In a Bayesian framework, they can also generate individual-level predictions together with their uncertainty, making them particularly well suited to the investigation of individual differences.
During the first two days, the workshop will cover the interpretation of random effects; individual-level predictions; model-based approaches to assessing whether observed individual differences can be measured reliably; and extensions in which variability may differ across groups or experimental conditions. Throughout the workshop, the brms R package will be used.
On the third day, participants will apply Bayesian hierarchical models to their own datasets and research questions or to a dataset provided by the instructor. The workshop will be mainly practical and interactive, combining presentations of the main principles with hands-on exercises completed individually and in groups. Resources will include presentation slides, datasets, exercises with annotated solutions, and a reading list.
Prior experience with R is required. Participants should also have a good grounding in regression modelling and some experience with mixed-effects models and Bayesian analysis. The workshop will be held in English and is limited to 20 participants. Priority will be given to members of GRADE Center Language and SFB 1629 NegLab.
Cancellations may be made up to one week before the workshop by sending an explanatory email. Failure to attend without prior notice will result in exclusion from one future GRADE Center Language workshop.