This is a list of resources that have been mentioned during the Statistical Rethinking course in places such as the online discussion, Discord chat, lectures, etc
| Name/Link | Description |
|---|---|
| Bayesian Workflow (Gelman et al 2020) | 100p paper on Bayesian workflow, Gelman is also working on a book version of this |
| Telling Stories with Data (Alexander) | post-stratification with code included, but also generally about statistical communication, programming, and modeling |
| Regression and Other Stories (Gelman, Hill, Vehtari 2020) | Site has a link to pdf of book |
| Active Statistics (Gelman, Vehtari 2024) | Book designed to accompany Regression and Other Stories, hundreds of stories, activities, and discussion problems on applied statistics and causal inference |
| Towards A Principled Bayesian Workflow (Betancourt 2020) | |
| Bayesian Statistics without Frequentist Language (McElreath, 2017) | Building a model slowly over time and then adding complexity |
| Name/Link | Description |
|---|---|
| What Is Your Estimand? Defining the Target Quantity Connects Statistical Evidence to Theory (Lundberg, Johnson, Stewart 2021) | Every quantitative study must be able to answer the question: what is your estimand? |
| Modeling Social Behavior: Mathematical and Agent-Based Models of Social Dynamics and Cultural Evolution (Smaldino 2023) (worldcat) | How to build agent based models and interrogate them, former McElreath student |
| Models of Social Dynamics | Intro lecture series on agent based modeling by by Paul Smaldino |
| Name/Link | Description |
|---|---|
| MCMC Interactive Gallery | Visualization of Hamiltonian Monte Carlo (HMC) |
| Bulk effective sample size (bulk-ESS) | ess_bulk as a new diagnostic replacing n_eff |
| Name/Link | Description |
|---|---|
| Nabiximols treatment efficiency - Improved LOO computation (Vehtari 2024) | Example of model comparison using non-ideal models. Generally, comparisons using models with known flaws useful for incremental model development, exploring what other models to try, and informative in and of itself |
| Projective inference in high-dimensional problems: Prediction and feature selection (Piironen, Paasiniemi, Vehtari 2020) | Procedure fitting a reference model and comparing with simpler models via algorithm that is less computationally intensive than refitting every model. Authors have a related projpred R package |
| Name/Link | Description |
|---|---|
| What's Wrong with Change In General? | Explanation of why in general change scores are a bad idea. A section of Biostatistics for Biomedical Research online book by Frank Harrell |
| Analyzing ordinal data with metric models: What could possibly go wrong? (Liddell, Kruschke 2018) | dangers of analyzing ordinal data such as likert scales with a metric model |
| How conditioning on post-treatment variables can ruin your experiment and what to do about it (Montgomery, Nyhan, Torres 2018) | Dangers of conditioning on post-treatment variables |
| Randomisation is not about balance, nor about homogeneity but about randomness | blog post about what randomization is for |
| Seven myths of randomisation in clinical trials | article by same author as above |
| Common Probability Problems Arise from Simple Invariances (Frank 2017) | technical |
| Name/Link | Description |
|---|---|
| Bayesian Inference without Probability Density Functions (Goodrich 2021) | Framework for using priors created by eliciting quantiles from subject matter experts |
| Quantile Function Stan Code | Using priors created by eliciting quantiles from subject matter experts, Ben Goodrich is someone that works on this |
| Hybrid elicitation and quantile-parametrized likelihood (Perepolkin, Goodrich, Sahlin 2023) | Using priors created by eliciting quantiles from subject matter experts |
| Name/Link | Description |
|---|---|
| Probability Theory as Logic (Jaynes 1990) | epistemological vs. ontological assumptions in statistics. Technical |
| Probability Theory: The Logic of Science (Jaynes 1995 posthumously?) | epistemological vs. ontological assumptions in statistics. Technical, rough, and weird in places. First few chapters relevant |
| Jaynes bibliography with pdfs |
| Name/Link | Description |
|---|---|
| Applied Longitudinal Data Analysis by Singer and Willett | Textbook with in-depth treatment but not a modern approach to estimation |
| Informed Bayesian Survival Analysis (Bartos, Aust, Haaf 2022) | Relevant BMC Medical Research Methodology article |
| Hierarchical Bayesian continuous time dynamic modeling (Driver, Voelkle 2018) | Continuous time dynamic model analysis approach |
| Hierarchical Continuous Time Dynamic Modelling for Psychology and the Social Sciences (Driver 2017) | Dissertation by same author as above (may be easier for beginner to read than above article) |
| Name/Link | Description |
|---|---|
| A guide to modeling proportions with Bayesian beta and zero-inflated beta regression models | Tutorial on modeling outcomes that range from 0-1 |
| Hierarchical Modeling (Betancourt, 2020) | technical, more about the funnel problem and multilevel models, examples of when centered vs non-centered models are better |
| Complex models and reparameterization - Box 11.2 Cholesky factorization | More in-depth explanation of how Cholesky factorization works |
| Entropy (for data science) Clearly Explained | Information entropy derivation explained as surprise |
| Name/Link | Description |
|---|---|
| Steve Frank papers | He has papers on which generative models correspond to which probability distributions |
| Regression Modeling Strategies (Harrel 2015 and 2024) | Chapter 2.4 tackles splines |
| Visualizing the differences between Bayesian posterior predictions, linear predictions, and the expectation of posterior predictions (Heiss, 2022) | |
| Prediction can be safely used as a proxy for explanation in causally consistent Bayesian generalized linear models (Scholz, Burkner 2022) | models under misspecification |
| White House "Nelson" memo | future requirement that publications and research data arising from US federal funds be made publicly accessible immediately upon publication |
| Bayesian Data Analysis (Gelman, Vehtari, et al 2021) | Gaussian processes don't scale well, so fast approximations are used. This book contains a derivation of approximations of Gaussian proccesses. Also has a birthday example of Gaussian processes |
| Bayesian workflow book - Birthdays (Vehtari 2020) | Case study page for birthday example above |
| Deep Neural Networks as Gaussian Processes (Sohl-Dickstein et al 2017) | link between Gaussian proesses and neural networks |
| Name/Link | Description |
|---|---|
| International Conference on Multilevel Analysis | |
| Elements of Evolutionary Anthropology | McElreath blog, "Occasional text on evolutionary anthropology, statistical inference, and the intersection of the two." |
| Stan forums | Online community for applied statistical models |