Bayesian analysis and Statistical Rethinking posts

1 minute read

I haven’t been shy about being a fan of Statistical Rethinking by Dr. Richard McElreath. Reading the book, following lectures, and doing problems, has helped me be more comfortable with statistical concepts that previously went over my head.

One of the things that helped me learn was writing and coding out solutions. These posts may have been problems directly from the book. Other times I went down rabbit holes stemming from my own curiosity. I’m compiling the posts here to help those who may be going through the course or learning PyMC. The order below is intended to be more logical for learning than my blog post page which shows posts in reverse chronological order.

Bayesian statistics-related posts before Statistical Rethinking

  1. An intro to Bayes’ theorem with Bertrand’s box paradox

  2. Bayes-ball part 1

  3. Bayes-ball part 2

  4. Bayes-ball part 3

Posts related to Statistical Rethinking

  1. Prior, prior predictive, likelihood, posterior, and posterior predictive]

  2. Running models forwards and backwards

  3. Correlated data, different DAGs

  4. Linear regression part 1: an intro to PyMC objects

  5. Linear regression part 2: understanding the posterior distribution

  6. Linear regression part 3: predicting an average outcome

  7. Linear regression part 4: predicting an actual outcome

  8. Multilevel modeling with binomial GLM

  9. Exploring multilevel modeling failure

  10. Escaping the Devil’s Funnel

  11. Weird ways that covariance matrices are made

  12. When the Spiderman meme is relevant to multilevel models

Of course, there’s still more for me to learn and these posts aren’t perfect. The image shown above is from a post on diagnosing model failure, which was in one of the later lessons of the book. I wanted to highlight it because “failure” is necessary for learning.

I’d welcome feedback at ben.lacar AT