Statistical Rethinking (2024 Edition) is an online course focused on scientific models in data analysis. Led by Richard McElreath, this course emphasizes causal models and Bayesian methods to derive insights from complex data. Join a community of learners and enhance your understanding of data through powerful computational tools.
Statistical Rethinking (2024 Edition)
Instructor
Richard McElreath
Course Overview
This course is designed to teach data analysis with a clear focus on scientific modeling. Understanding that data by itself cannot tell a complete story, we emphasize the importance of expressing what caused the data. Learners will prioritize conceptual and causal models while formulating precise questions correlated to these frameworks. The course leverages Bayesian data analysis to connect scientific models to evidence, equipping students with robust computational tools to handle high-dimensional, imperfect data typical in biostatistics and social sciences.
Format
The course will be delivered online using a flipped instructional model. Lectures will be pre-recorded and released twice a week, complemented by weekly online discussion sessions every Friday from 3 PM to 4 PM Central European Time (Berlin). This schedule is designed to be accommodating for participants from the Americas.
Students will utilize the 2nd edition of the book Statistical Rethinking and may also engage with draft chapters for the upcoming 3rd edition, which will be provided in PDF format to enrolled students.
Registration
Unfortunately, this year's registration reached its cap of 200 participants, and the waitlist is currently full. Should any registered students withdraw or stop attending discussions, we will invite individuals from the waitlist. Details can be found in the Wait List.
Course Calendar & Topical Outline
Over the span of 10 weeks, students will engage with lecture recordings and complete weekly problem sets assigned each Friday. Below is the topical outline of the course:
Coding Component
The course has a robust coding component, allowing students to practice the material through practical applications. Participants can choose from different platforms:
- Original R: Utilize R code examples from the print book. Install the
rethinkingR package, available on GitHub for more details, including the option of usingcmdstanras an updated MCMC engine. - R + Tidyverse + ggplot2 + brms: A comprehensive and high-quality conversion available through Tidyverse/brms.
- Python and PyMC3: Find a complete conversion at the Python/PyMC3 repository.
- Julia and Turing: Although not as comprehensive, the Julia/Turing conversion presents a growing assortment of Rethinking examples across the Julia language.
For a detailed list of additional resources, visit: Statistical Rethinking Resources.
Homework and Solutions
Weekly problem sets and their solutions will be shared throughout the course, accessible in the designated folders at the top of the repository. Engage with the course material thoughtfully and improve your data analysis skills today!