Here’s a full draft of the upcoming second edition of my “Data Visualization: A Practical Introduction”: socviz.co
05.03.2026 22:54
👍 553
🔁 175
💬 13
📌 15
Here’s a full draft of the upcoming second edition of my “Data Visualization: A Practical Introduction”: socviz.co
Toward the end of his two-volume Treatise on the Venom of the Viper, published in 1781, the Tuscan naturalist Deluxe Fontana declared: "I have made more than 6000 experiments; I have had more than 4000 animals bit; I have employed upwards of 3000 vipers and may have been deceived; some essential circumstance may have escaped me: I may have neglected some other, not thinking it necessary; my consequences may have been too general, my experiments too few in number. In a word, I may very easily have been mistaken, and it would be almost impossible that I should never have been so in a matter so difficult, so obscure, and likewise so new."
Fontana thinking his 6000+ experiments may not have been enough to feel confident in his conclusions in the 18th century while we expect our singular experiments and their standalone replications to do wonders some 250 years later... (excerpt from the intro of Jutta Schickore's About Experiment)
![2. Install the diagram extension in a Quarto project. Create a new Quarto project. From the terminal, run this to install the diagram extension:
quarto install extension pandoc-ext/diagram
3. Tell your document to use the extension. Add this to the YAML header of a Quarto document:
filters:
- diagram
4. Add a .tikz block to your document. Add a tikz code block to your document like this:
```{.tikz}
\begin{tikzpicture}[>={stealth}]
\node (x) at (0,0) {$X_{t}$};
\node (y) at (2,0) {$Y_{t + 1}$};
\node (z) at (1,1) {$Z$};
\path[->] (z) edge (x);
\path[->] (z) edge (y);
\path[->] (x) edge (y);
\end{tikzpicture}
```
5. Render! Render the document to PDF and you should see a diagram. Render it to HTML or .docx or Typst and it should use inkscape to convert it to SVG. Magic.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:2zcfjzyocp6kapg6jc4eacok/bafkreifut33lozgqgp3ualtpkyze47rtkjoysfginsohfykakfibqzetyu)
2. Install the diagram extension in a Quarto project. Create a new Quarto project. From the terminal, run this to install the diagram extension: quarto install extension pandoc-ext/diagram 3. Tell your document to use the extension. Add this to the YAML header of a Quarto document: filters: - diagram 4. Add a .tikz block to your document. Add a tikz code block to your document like this: ```{.tikz} \begin{tikzpicture}[>={stealth}] \node (x) at (0,0) {$X_{t}$}; \node (y) at (2,0) {$Y_{t + 1}$}; \node (z) at (1,1) {$Z$}; \path[->] (z) edge (x); \path[->] (z) edge (y); \path[->] (x) edge (y); \end{tikzpicture} ``` 5. Render! Render the document to PDF and you should see a diagram. Render it to HTML or .docx or Typst and it should use inkscape to convert it to SVG. Magic.
New blog post! After years of annoying shenanigans, I discovered that nowadays it's *really* straightforward to get tikz diagrams in #QuartoPub to automatically turn into SVGs in HTML, docx, and Typst! Install Inkscape, install a Quarto extension, and you're done. www.andrewheiss.com/blog/2026/02...
course schedule as a table. Available at the link in the post.
I'm teaching Statistical Rethinking again starting Jan 2026. This time with live lectures, divided into Beginner and Experienced sections. Will be a lot more work for me, but I hope much better for students.
I will record lectures & all will be found at this link: github.com/rmcelreath/s...
If you have to read anything about the prospect of “automating scientific discovery,” “agents for science,” or integrating LLMs into scientific pipelines, please let it be this essay by Kevin T. Baker: artificialbureaucracy.substack.com/p/context-wi...
A blog post giving a more thorough take on survey experiments and the credibility revolution: cyrussamii.com?p=4168
DAG display
Rearranging variables can make this clearer. In plot below, it's relatively easy to see that total effect needs no adjustment, and that direct effect from A to Y needs to block the indirect path that traverses D and C. D is a collider, and C the child of the collider, hence adjustment on B is needed
"Is it really impossible to imagine productive engagement with sociologists who take Bourdieu or Bauman seriously?"
Well
https://poorlydrawnlines.com/
poorlydrawnlines.com
Ewige Wiederkunft but for academics.
Great piece on the absurdity of brute force multiverse analyses.
www.pnas.org/doi/10.1073/...
A meme with two panels. In panel 1 the woman looks disgusted. In panel 2 intrigued. Text, panel 1: fake data. Text, panel 2: silicon sample
Preprint and the website by @dingdingpeng.the100.ci and @vincentab.bsky.social are great, and I got inspired to illustrate Bayesian workflow for model checking and comparison *before* model interpretation following their friendship importance example users.aalto.fi/~ave/casestu...
Models as Prediction Machines: How to Convert Confusing Coefficients into Clear Quantities Abstract Psychological researchers usually make sense of regression models by interpreting coefficient estimates directly. This works well enough for simple linear models, but is more challenging for more complex models with, for example, categorical variables, interactions, non-linearities, and hierarchical structures. Here, we introduce an alternative approach to making sense of statistical models. The central idea is to abstract away from the mechanics of estimation, and to treat models as “counterfactual prediction machines,” which are subsequently queried to estimate quantities and conduct tests that matter substantively. This workflow is model-agnostic; it can be applied in a consistent fashion to draw causal or descriptive inference from a wide range of models. We illustrate how to implement this workflow with the marginaleffects package, which supports over 100 different classes of models in R and Python, and present two worked examples. These examples show how the workflow can be applied across designs (e.g., observational study, randomized experiment) to answer different research questions (e.g., associations, causal effects, effect heterogeneity) while facing various challenges (e.g., controlling for confounders in a flexible manner, modelling ordinal outcomes, and interpreting non-linear models).
Figure illustrating model predictions. On the X-axis the predictor, annual gross income in Euro. On the Y-axis the outcome, predicted life satisfaction. A solid line marks the curve of predictions on which individual data points are marked as model-implied outcomes at incomes of interest. Comparing two such predictions gives us a comparison. We can also fit a tangent to the line of predictions, which illustrates the slope at any given point of the curve.
A figure illustrating various ways to include age as a predictor in a model. On the x-axis age (predictor), on the y-axis the outcome (model-implied importance of friends, including confidence intervals). Illustrated are 1. age as a categorical predictor, resultings in the predictions bouncing around a lot with wide confidence intervals 2. age as a linear predictor, which forces a straight line through the data points that has a very tight confidence band and 3. age splines, which lies somewhere in between as it smoothly follows the data but has more uncertainty than the straight line.
Ever stared at a table of regression coefficients & wondered what you're doing with your life?
Very excited to share this gentle introduction to another way of making sense of statistical models (w @vincentab.bsky.social)
Preprint: doi.org/10.31234/osf...
Website: j-rohrer.github.io/marginal-psy...
Ah, the author is "that" guy.
I am confused. Your title mentions multinomial independent variables, and then you discuss "questions about different experiences in the healthcare system" (implying multiple questions), and then "this variable is dichotomized" sounds like a single variable, and nothing multinomial about that. Can you please explain just what your indepedent and dependent variables are? (Also, can you move away from dichotomania?) – Stephan Kolassa Commented2 hours ago @Stephan Kolassa The problem is misnamed. The problem is dichotophilia, an unreasonable affection for dichotomies. – Nick Cox Commented2 hours ago @NickCox: I have to respectfully disagree. Mania is characterized by "a state of abnormally elevated arousal, affect, and energy level", which quite evidently is involved when people apparently derive deep satisfaction from dichotomizing inherently numerical observations. – Stephan Kolassa Commented2 hours ago Why do we say necrophilia? (Actually I don't usually talk about it, but there you go.) – Nick Cox Commented1 hour ago https://stats.stackexchange.com/questions/668701/what-is-the-right-way-to-handle-multinomial-independent-variables-in-logistic-re
One more reason to be on CV.
New #rstats blog up!
solomonkurz.netlify.app/blog/2025-07...
This is the first in a brief series where we use {brms} to learn {Stan} code.
Many thanks to @fusaroli.bsky.social and @stephenjwild.bsky.social for their helpful reviews.
{tinytable} is a dead simple, ultra-flexible, and dependency-free #Rstats 📦 to turn data frames into beautiful tables: html, word, pdf, latex, typst, markdown, etc.
v0.10.0 has cool new features and important bug fixes. Check out the detailed tutorials at:
vincentarelbundock.github.io/tinytable/
Oh, no: www.york.ac.uk/depts/maths/...
Every time I write "the literature is mixed" what I mean is "I don't want to talk about it"
I don't know, but my guess is this: successful theories in the life sciences (e.g., Darwinian evolution, learning theory, economics, Downs' 1957 theory of party competition) center on attempts by an actor to maximize some single something (gene survival, rewards, utility, votes). But groups, other than commodity traders and political parties, either maximize many things or just sit there and don't maximize anything. And if we downshift from groups to individuals, it is unlikely we can beat psychology or economics at the individual actor basic theory game.
This also reminds me of Davis' argument (I know you know):
If I let people around me to decide on this shared theory, I have to use terms like habitus, field, and capital culturel only. I understand the allure of something like utility model, but I don't think something similar would work for sociology, at least not now.
I agree with your main argument and critique of so-called heterodox people (e.g., procedural prescriptions, nothing of substance). I am not sure about “the boring reasons”, especially the first one.
New Substack post
It's very personal: my story of a 20-year academic career, and the many challenges of theoretical and cross-disciplinary work
As I put it in the subtitle: There is a lot of success and a lot of pain here, and no happy ending
thomscottphillips.substack.com/p/happy-in-t...
