Introduction to Bayesian data analysis

IMPRS NeuroCom

  • Start: Mar 10, 2023 10:30 AM (Local Time Germany)
  • End: Mar 16, 2023 05:30 PM
  • Speaker: Prof. Dr Daniel Schad
  • Location: Max Planck Institute for Human Cognitive and Brain Sciences
  • Host: IMPRS Coordination
  • Contact: imprs-neurocom@cbs.mpg.de

Friday, March 10
10:30 - 12:00 Basic Concepts
12:00 - 13:00 Lunch break
13:00 - 14:30 Introduction to Bayes
14:30 - 15:00 Coffee break
15:00 - 16:30 Bayesian linear models 1

Tuesday, March 14
10:30 - 12:00 Bayesian linear models 2
12:00 - 13:00 Lunch break
13:00 - 14:30 Exercise linear models
14:30 - 15:00 Coffee break
15:00 - 16:30 Hierarchical Models

Thursday, March 16
10:30 - 12:00 Exercise hierarchical models
12:00 - 13:00 Lunch break
13:00 - 14:30 Model comparison
14:30 - 15:00 Coffee break
15:00 - 16:30 Exercise model comparison

Potential additional topics: Principled Bayesian workflow + Bayes factor workflow

Content:
This three-day workshop provides an introduction to concepts of Bayesian analysis, to Bayesian linear and linear mixed-effects models, as well as to Bayes factor null hypothesis tests. It discusses how to conduct such analyses in the cognitive sciences, and shows how to code example analyses in R (using brms). The course assumes knowledge of frequentist statistics, including linear (mixed-effects) models (in lme4), as well as familiarity with R.

About the instructor:
Daniel Schad is professor for Quantitative Methods at the HMU Health and Medical University in Potsdam. Before, he was assistant professor for Cognitive Science and Artificial Intelligence at University of Tilburg (Netherlands) and held research positions at the Charité - Universitätsmedizin Berlin and at the University of Potsdam, where he also got his PhD in cognitive and mathematical psychology.

Recommended reading:
Nicenboim, B., Schad, D.J. & Vasishth, S. (2023). An introduction to Bayesian data analysis for cognitive science. https://vasishth.github.io/bayescogsci/book/

Software:
Please install R and R-Studio on your computer. Moreover, the following R-packages should be installed: rstan, brms, tidyverse, bcogsci, extraDistr, bridgesampling, bayesplot, MASS, designr. Some of these installations are not straightforward and may take some time - I recommend to do this in time before the course starts. Additional packages may need to be installed during the workshop.

Target group: Doctoral researchers of IMPRS NeuroCom and other MPI CBS doctoral researchers

Awarded ECTS: 0.5 ECTS (80% attendance and in-course assignments)

 

 

Go to Editor View