Vincent Arel-Bundock
LightNews LightNews
banner
vincentab.bsky.social
Vincent Arel-Bundock
@vincentab.bsky.social
6.2K followers 1.4K following 950 posts
Prof. Most tweets about R. “Polisci, it’s all about what’s going on.”

http://arelbundock.com
Posts Replies Media Videos
Pinned
vincentab.bsky.social
Vincent Arel-Bundock @vincentab.bsky.social · Sep 17
Whoa—my book is up for pre-order!

𝐌𝐨𝐝𝐞𝐥 𝐭𝐨 𝐌𝐞𝐚𝐧𝐢𝐧𝐠: 𝐇𝐨𝐰 𝐭𝐨 𝐈𝐧𝐭𝐞𝐫𝐩𝐫𝐞𝐭 𝐒𝐭𝐚𝐭 & 𝐌𝐋 𝐌𝐨𝐝𝐞𝐥𝐬 𝐢𝐧 #Rstats 𝐚𝐧𝐝 #PyData

The book presents an ultra-simple and powerful workflow to make sense of ± any model you fit

The web version will stay free forever and my proceeds go to charity.

tinyurl.com/4fk56fc8
vincentab.bsky.social
J'adore les chaises à droite. Funky!
February 11, 2026 at 1:01 PM
vincentab.bsky.social
looks great!
February 4, 2026 at 1:23 PM
Reposted by Vincent Arel-Bundock
dingdingpeng.the100.ci
New preprint! So, what's a multiverse analysis good for anyway?>

With @jessicahullman.bsky.social and @statmodeling.bsky.social

juliarohrer.com/wp-content/u...
What’s a multiverse good for anyway? Julia M. Rohrer, Jessica Hullman, and Andrew Gelman Multiverse analysis has become a fairly popular approach, as indicated by the present special issue on the matter. Here, we take one step back and ask why one would conduct a multiverse analysis in the first place. We discuss various ways in which a multiverse may be employed – as a tool for reflection and critique, as a persuasive tool, as a serious inferential tool – as well as potential problems that arise depending on the specific purpose. For example, it fails as a persuasive tool when researchers disagree about which variations should be included in the analysis, and it fails as a serious inferential tool when the included analyses do not target a coherent estimand. Then, we take yet another step back and ask what the multiverse discourse has been good for and whether any broader lessons can be drawn. Ultimately, we conclude that the multiverse does remain a valuable tool; however, we urge against taking it too seriously.
February 4, 2026 at 10:24 AM
Reposted by Vincent Arel-Bundock
gsimpson.bsky.social
🚨 GAMs have moved on—so it’s time for an update.

On March 3, 2026 (17:00–19:00 CET) I’ll be livestreaming an updated introduction to Generalized Additive Models in R

📺 YouTube livestream link: youtube.com/live/A9U8e1K...

#RStats #mgcv #GAMs #gratia #statistics 🧪
What's new in the world of Generalized Additive Models
YouTube video by Bottom of the Heap
youtube.com
February 3, 2026 at 12:57 PM
vincentab.bsky.social
I know you prefaced with: "if the slope is significant." I'm saying that's not always the case, so (a) there's a high risk these plots are misleading, (b) they focus our eyes on unimportant features, and (c) add basically no new information.
February 2, 2026 at 1:01 PM
vincentab.bsky.social
Counterpoint: they two points would be colored differently on a J-N plot but they are indistinguishable.
February 2, 2026 at 12:59 PM
vincentab.bsky.social
Hard disagree github.com/vincentarelb...
February 1, 2026 at 9:44 PM
Reposted by Vincent Arel-Bundock
paulgp.com
One of the most important statistical packages made in Econ in the last decade
paperposterbot.bsky.social ArXiv Paperboy (Stat.ME+Econ.EM) @paperposterbot.bsky.social · 11d
arXiv📈🤖
Fast and user-friendly econometrics estimations: The R package fixest
By Berg\'e, Butts, McDermott
fixest is an R package for fast and flexible econometric estimation, providing a comprehensive toolkit for applied researchers. The package particularly excels at fixed-effects estimation, supported by a novel fixed-point acceleration algorithm implemented in C++. This algorithm achieves rapid convergence across a broad class of data contexts and further enables estimation of complex models, including those with varying slopes, in a highly efficient manner. Beyond computational speed, fixest provides a unified syntax for a wide variety of models: ordinary least squares, instrumental variables, generalized linear models, maximum likelihood, and difference-in-differences estimators. An expressive formula interface enables multiple estimations, stepwise regressions, and variable interpolation in a single call, while users can make on-the-fly inference adjustments using a variety of built-in robust standard errors. Finally, fixest provides methods for publication-ready regression tables and coefficient plots. Benchmarks against leading alternatives in R, Python, and Julia demonstrate best-in-class performance, and the paper includes many worked examples illustrating the core functionality.
January 30, 2026 at 5:03 PM
vincentab.bsky.social
I wish I could speed more things than just code production. Like, say, IRB approval.
January 29, 2026 at 3:10 PM
vincentab.bsky.social
lol
January 29, 2026 at 2:48 PM
Reposted by Vincent Arel-Bundock
gmcd.bsky.social
Things are grim. But in more frivolous news...

@jamesbrandecon.bsky.social and I have been chipping away at `dbreg`, a 📦 for running big regression models on database backends. For the right kinds of problems, the speed-ups are near magical.

Website: grantmcdermott.com/dbreg/

#rstats

[1/2]
dbreg
grantmcdermott.com
January 26, 2026 at 4:57 PM
vincentab.bsky.social
FYI, broom-ified shortcuts: `modelsummary::get_gof()` and `modelsummary::get_estimates()`. Obviously, if your students use other `easystats` stuff, it's worth introducing them to the ecosystem, especially since they are doing the actual "work" here.
January 23, 2026 at 1:04 AM
Reposted by Vincent Arel-Bundock
mattansb.msbstats.info
Finished teaching my new Advance Stats for Psych graduate course today with a heavy emphasis on both DAGs and shifting away from coefficient interpretation and towards models as prediction machines.

Both went great!

The latter was extremely helpful for logistic regression (for obvious reasons 😵‍💫)!
dingdingpeng.the100.ci Julia M. Rohrer @dingdingpeng.the100.ci · Aug 25
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...
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.
January 19, 2026 at 6:28 PM
vincentab.bsky.social
Thanks, that's very useful and interesting!
January 14, 2026 at 2:58 AM
vincentab.bsky.social
hmm, can't think of a direct solution now, but it sounds pretty easy to achieve this by manipulating the draws a bit. I posted this minimal "solution" to the forum.
January 14, 2026 at 1:34 AM
vincentab.bsky.social
Yes, in principle, I think that should work. You are just missing: c()
January 14, 2026 at 12:51 AM
vincentab.bsky.social
My poster got allocated prime real estate at the conference: High traffic, captive audience.
January 6, 2026 at 1:19 PM
Reposted by Vincent Arel-Bundock
s3alfisc.bsky.social
There is a new-ish Python package, moderndid, that implements almost all of the "modern" DiD estimators:
github.com/jordandekler...
GitHub - jordandeklerk/moderndid: Python package implementing modern DiD estimators with diagnostic tools and sensitivity analysis.
Python package implementing modern DiD estimators with diagnostic tools and sensitivity analysis. - jordandeklerk/moderndid
github.com
January 3, 2026 at 10:40 AM
vincentab.bsky.social
very exciting stuff. Congrats!
December 24, 2025 at 8:05 PM
Reposted by Vincent Arel-Bundock
michaelchirico.bsky.social
Look under your tree! 🌲 🎁

There's a major #rstats #rdatatable release waiting!

A tremendous thanks to all involved and especially those contributing to some major (performance-maintaining!!) rewrites around the non-API issue.

cran.r-project.org/web/packages...
December 24, 2025 at 7:00 PM
vincentab.bsky.social
looks great!
December 23, 2025 at 2:27 PM
vincentab.bsky.social
Oh yeah, I'm all in for validation and code review. Essential!
December 22, 2025 at 6:29 PM
vincentab.bsky.social
I mainly use claude and codex from the command line, and gpt web interface.
December 22, 2025 at 4:45 PM
vincentab.bsky.social
Have you tried other similar tools? Copilot is the one I use the least among them. FWIW, care about "correct" code, I code a lot, and can't imagine doing it without these tools anymore. (Not a fan of the dismissive language in the original post.)
December 22, 2025 at 4:41 PM
vincentab.bsky.social
I'm hoping they'll become better at continuous update. also, the new "Skills" feature in Claude (and similar elsewhere) might be useful. Developers could package docs, prompts, and instructions to guide LLMs.
December 21, 2025 at 3:09 PM