Announcing my forthcoming short paper at #ieeevis! In this solo-author work, I examine and critique the logic of generalization about decision support in visualization research. arxiv.org/abs/2508.06751

I'm extremely excited to share this new paper out of the Data Cognition Lab! In it @krisha-mehta.bsky.social, Gordon Kindlmann, and I reframe visualization as a mechanism for data disclosure and develop a vocabulary for how visualization design induces loss on underlying data signals.

Today, along with 2,000 other NIH employees, I had to clear out my office 😭 It was truly the honor of my life to work with such incredibly passionate people focused on improving human health. I’ve never experienced a more positive culture where *everyone* cared about their job and serving others.

I wish more HCI research was this careful about measurement. If we do turn the corner, I think we’re gonna find out that a lot of our study results don’t mean what the authors originally said they meant.

Thanks to the wonderful folks at @dsi-uchicago.bsky.social for making this video!

What tech company will be the first to stand up and say that gutting the NSF is bad actually? The economic argument seems pretty clear Less NSF --> fewer PhD students --> fewer researchers --> smaller AI tech pipeline --> slower progress --> less competitive globally

Most of your life won’t go as planned. Stop placing your happiness, health, and self-actualization on the horizon.

“The executive orders make it quite clear that certain DEI activities are considered to violate current law, and we cannot fund anything that includes activities deemed unlawful.” Okay, but until a few weeks ago, promises about DEI were required. Idk how any of our grants survive this litmus test.

Somewhere in the U.S., there’s a scientist staring at their NSF/NIH grant application wondering why they bother. This post is for you. Science and society both need you. Hang in there and know there is a whole community supporting you.

A great thread if you’re into visualization for model interpretability and Bayesian stats!

Anyone know how to turn off the AI assistance feature in Overleaf? I didn’t want this. I didn’t consent to this. I find the suggestions distracting in a way that basically makes the tool unusable for deep work. Am I alone in feeling this way?

This is so upsetting. I have two grants out for review, basically my whole summer worth of effort scuttled. Multiply this by thousands of scientists whose time and ideas are being thrown away. It’s especially concerning for early career researchers whose futures depend on securing funding.

I’m feeling for the researchers who are losing access to funding.

“Universal fruitstuff magnitudes” is now a term of art

Professors, don't be mean to job candidates, they remember it the rest of their lives. Source: remembered it the rest of my life.

📊 #dataviz #map Very cool interactive Voronoi diagram tessellation of the globe partitioned by nearest major airport www.jasondavies.com/maps/voronoi...

Just because I keep seeing the "replications show which effects are real" or "effects that didn't replicate once are flukes" heuristics repeatedly showing up on my TL today, I'm re-upping this thread. I beg everyone to please stop searching for quick and conclusive judgments on what's "true".

The distinction @jessicahullman.bsky.social makes between 1st and 2nd order uncertainty is important to understanding prediction vs inference. Conditional predictive performance (1st order) can be checked, but whether we’ve learned the right distribution (2nd order) is what we need to generalize.

This made me laugh.

How wrongheaded to think that “engagement” without communicative intent will keep actual people on a dying platform. Meta is phoning it in at this point. This bit about appealing to younger audiences is either wishful thinking or a blatant lie to assuage nervous investors.

Check out the earthquakes over the past year with this interactive globe: observablehq.com/@joewdavies/... You can toggle between country borders and tectonic plates. I quite like being able to gauge the depth of them. #earthquakes #seismology #cartography #globe #dataviz

“Where are the findings? You only provide quotes.” New article considers qualitative researchers’ experiences of methodologically incongruent peer review feedback Open Access: dx.doi.org/10.1037/qup0... Few quotes follow 🧵

This guy chased me off of the psych methods Facebook group about 10 years ago. I’m not surprised to hear others had issues with him. You really don’t want to be remembered by your colleagues primarily for verbal abuse. What an unfortunate way to live.

Writing is thinking. It’s not a part of the process that can be skipped; it’s the entire point.

I imagine this is how most statistical consulting turns out. There’s only so much you can do with “garbage in.”

Super cool way to move between aggregate and discrete distributional representations! Reminds me of Steve Drucker’s work.

My colleague Rana Hanocka does exciting work on the frontier for 3D graphics and AI/ML, and she’s recruiting PhD students this cycle!

Has anyone written about the privacy externalities of the decision to use a service like 23andMe? Users are sharing data not only about their own genome but also anyone related to them. This seems a little tricky from the perspective of consent and data governance 🤔

I’m on a real Paramore kick right now 🎸

UChicago also probably inflated by MAs here. The Chetty et. al. results for undergrad admissions show much less preference for top 1% parental income.

This panel today was great! It’s nice to have candid conversations with students about growing as a researcher. Thanks to @tengshanyuan.info and @krisha-mehta.bsky.social for organizing!

I already weighed in, but do others have recommendations?

Nice new website for @puddingviz.bsky.social pudding.cool Quite a body of work too! #dataviz

Please volunteer for the VIS 2025 PC!

I don't agree that the second option is the solution unless we correctly diagnose what the problem is. Replicability is a red herring. We can fix that artificially without making any real progress toward better knowledge/explanations/understanding. You can have very robust false results.

I am recruiting a postdoc! The research emphasis is participatory design for decision-making interfaces. Check out the two project descriptions on my website. people.cs.uchicago.edu/~kalea/prosp... The position is through UChicago’s Data Science Institute. datascience.uchicago.edu/research/pos...

My best friend put out an album this weekend after at 10+ year hiatus. It’s a vulnerable and beautiful account of breaking through self-doubt, social isolation, and alcohol abuse. Give it a listen if you like indie and R&B, or if you’re longing for something hopeful open.spotify.com/album/7c6kRv...

A statement I made in the talk, with a hat tip to a well known quote/title: The Difference Between Reproducible and Not Reproducible is not Itself Reproducible. I'd like to think Gelman and Stern wouldn't mind the rephrasing

cope with uncertainty by becoming one with uncertainty 🌟

I like this way of presenting ordinal data! I think people are sometimes too quick to use stacked bar charts for ordinal data when histograms can be simpler and easier to read #datavis #rstats

This is among the best internship opportunities out there imo. Amazing team of people!

Is anyone else just not paying any attention to the polls and election forecasts this year? All that data collection, modeling, and interactive vis just to say it’s close.

Entire industries and research programs are effectively chasing a marketing campaign. When did we start deciding what problems are important to work on this way? I am torn between thinking that LLMs are an important technology to study and that this is an unscientific way to prioritize R&D.

We were honored to host the IEEE VIS Conference Chicago Satellite event (organized by @kalealex.bsky.social), celebrating contributions to visualization research and practice!

📚🚨 I posted 11 new chapters of my upcoming book! Model to Meaning: How to Interpret Statistical Results with #marginaleffects for #RStats and Python. These are early drafts and I really need your feedback! Errors, content requests, improvements, etc. marginaleffects.com

The folks at @dsi-uchicago.bsky.social were gracious to host our IEEE VIS 2024 Satellite event! They helped my students and I plan this event—which brought together graduate students and faculty from UChicago, DePaul, UIC, and Northwestern—with less than a week’s notice.

UpSet won the #ieeevis InfoVis 10-year test of time award 🎉. Thanks to @ngehlenborg.bsky.social @henstr.bsky.social Romain Vuillemot, and Hanspeter Pfister for being an awesome team to work with on this. I wrote up my reflections on what made UpSet successful: vdl.sci.utah.edu/blog/2024/10...