John Sakaluk

Very interesting/exciting (and congrats to y'all's team)!

Perhaps a dumb q: this looks great for data science purposes (and will prob. be my default suggestion to my students), but any idea how it fares for development?

Currently using (and happy w/) Cursor for 📦 stuff, but remain open to alts.

A screenshot of an RStudio window. On the left-hand side is a new pain called Posit Assistant. The Posit Assistant had recently run code making a lat-lon plot of Washington state, colored by whether the point had been marked as forested or not.

Today we're releasing AI for RStudio. It's really, really good—I'd encourage you to point it at the messiest data sources you have and see what it can do.

www.simonpcouch.com/blog/2026-03...

Bsky friends: if you get only one reading for a grad seminar on ethics in data analysis and reporting, what do you choose? @drandreahoward.bsky.social @kevinmking.bsky.social @foswald.bsky.social @marcuscrede.bsky.social @dingdingpeng.the100.ci @bwroberts.bsky.social @tedmond.bsky.social

Feel like it's difficult to "cover all bases" in one reading/one might need to choose a theme to emphasize and others to subordinate.

I'd probably sacrifice AI-theme (so rapidly unfolding), among others, for something QuantCrit-y. Castillo & Gillborn (2025) or Garcia et al. (2018) perhaps?

#rstats RIP, John Fox
@yihui.org just published this lovely tribute to John Fox and his work

yihui.org/en/2026/02/j...

- real sus that you want to require specific ID to vote and then immediately set precedent to instantaneously invalidate people’s ids

- this also includes a “bounty” on trans people using the bathroom. Which is so disgusting I cannot even express how bad of an idea this is

This is deeply alarming.

Absolutely appalling from the State I called home for 5 years

Those were good times, and pretty sure y'all were the only ones I was ever able to trade with for Gengars and Dragonites... I still have a pokemon-themed GB color + red cartridge that I'll play through every couple of years

Splitting the $ of a blue cartridge with Colin and Alain in grade 4...couldn't wait for my crack at the game.

So hard to choose, so I went with... 1) my fav. evolved starter, 2) my long-term/grind choice; and 3) I always had to have one, just because

YouTube video by Movies Déjà Vu Amadeus (1984) - Mozart Laughing Selection

Take care, as the "hahahahaha" {hahahahahavaan} is pronounced the Tom Hulce way:

www.youtube.com/watch?v=8PWb...

Quick, someone make the {hahahahavaan} package!

Paper on statistical power necessary for interaction effects
doi.org/10.1177/2515...

it immediately squashed it.

IT WAS/IS THE PERFECT EXAMPLE for when I return to teaching of why you still need to know your shit when deploying AI technologies (and how with oversight, they can offer tremendous value)!

...and so I got the AI to generate me a summary of what happened to share 😈

Friends, I just used an AI-tool in an R 📦 dev task; it found errors, and insisted it was a nothing-burger/probably dealing with an unrelated package; I reviewed what the AI-created functionality returned, and noticed a key, subtle omission; I pointed it to this problem; and >

adrftools: Estimating, Visualizing, and Testing Average Dose-Response Functions Facilitates estimating, visualizing, and testing average dose-response functions (ADRFs) for characterizing the causal effect of a continuous (i.e., non-discrete) treatment or exposure. Includes suppo...

I'm so excited to announce the first release of my newest #Rstats package, {adrftools}! This package facilitates estimation, visualization, and testing for the causal effect of a continuous (i.e., non-discrete) treatment.

🧵 1/10

#statssky #episky #causalinference

Psych-y friends, anyone have any recommended readings for latest/greatest theories for labeling-related processes of interpersonal events and/or identities?

You should be freaking out about AI It works and it keeps getting better at an insane pace

"Even if I refuse to use AI for the supposed purposes of personal gratification that I get from reading and writing (manually), I’m not sure that’ll still seem worthwhile and enjoyable to me!

statsandsociety.substack.com/p/you-should...

For sure, and I'm lucky enough to develop in an area where the sensible "human heat-check" things to review/test are pretty obvious and straightforward.

But it took awhile--at least for me--to crest the psychological hump of figuring out my comfort with using it.

To wit: as a sabbatical stretch my brain activity, I am vibe-coding a mobile game (and as the author of that post described, "programming mostly in English").

I am an Okay R-programmer; I am a dog-shit C programmer. And yet, what I've cobbled together in a couple weeks is mind-numbing (for a noob)

are not really paying attention, or are swept up in some of the (understandable) first-order consequences (e.g., workload spiking through increased plagiarism cases) to stay connected to the rate of change with this tech

It took me so, so, so long to find--and I had to go outside of academe to a very code-savvy friend of a friend who works for a FANG company--what I felt was a 'balanced take' between the doomers and the [immediate] hype-monsters.

Have achieved some existential calm, but feels like most colleagues>

... then that's better "value" from me for the taxpayer and I ought to do it." So I've made my peace with involving it.

And while I'm somewhat sympathetic to the "intrinsic enjoyment" argument, as my uncle likes to quip, "If it was 'fun', it wouldn't be called 'work'".

I had this realization a few months back working on my R 📦 , where I'm increasingly using AI-bosted dev tools. I wondered whether I should be using them at all, or maintaining my "from scratch" approach. Landed on, "this [research] is my job, and if this tool helps me propel the software further >

Can you give an example of what such unit tests would look like? I might have too much "developer brain" at the moment, but I'm not intuiting what these tests would look like/offer beyond:
A) using a fully reproducible document type (e.g., .Qmd), and
B) rigorous unit testing in constituent pkgs

Model to Meaning: How to Interpret Statistical Models with R and Python Published in Journal of the American Statistical Association (Just accepted, 2026)

"Model to Meaning" just received a super generous review in JASA.

Check it out! (I'm blushing 😊)

Reminder: You can get the paper copy from CRC, and the full **free** version will remain online forever at marginaleffects.com

www.tandfonline.com/doi/full/10....

This approach is especially useful for #sexuality and #gender questions that use choose-all-that-apply options to better reflect people’s lived experiences (particularly in #trans research). I walk through the data-wrangling process with these examples variables!

📦: knickodem.github.io/CATAcode/

Using CATAcode for Choose-All-That-Apply Sexual Orientation and Gender Questions In mid-2025, I collected data from trans participants, and some of my survey questions were choose-all-that-apply. Since that time, Merrin and Nickodem (2025) released an R package, CATAcode, to he…

Started a blog! One of my first posts walks through how to prep your data to use the CATAcode #rpackage (with code + diagrams). With CATAcode, you can easily summarize and visualize responses from choose-all-that-apply questions:

catchingsplinters.wordpress.com/2026/01/27/u...

🧵 1/2

I’ve been thinking about how to find R packages by functionality when you don’t already know the package name.

So over the holidays, Claude Code and I built The Warehouse: a functionality-first R package directory that helps you find packages by what they do.
rwarehouse.netlify.app

#rstats

So all colleges and universities will bring back the DEI stuff they cut when obeying in advance right...............................................................................................................................................right?