Stephen Wild
@stephenjwild.bsky.social
I try to put straight lines through things but usually fail. Try to be Bayesian when I can. Views my own. RT/like != endorsement.
I'm so happy to announce version 2.0.0 of my #Rstats package WeightIt is out on CRAN! New features: censoring weights, multilevel propensity scores, improved weights for continuous treatments, bias-reduced ordinal and multinomial models, M-estimation in subgroups Check out the website below!
Weighting for Covariate Balance in Observational Studies
Generates balancing weights for causal effect estimation in observational studies with binary, multi-category, or continuous point or longitudinal treatments by easing and extending the functionality ...
ngreifer.github.io
In Manchester's biography of MacArthur, MacArthur comes off as an egotistical asshole. Even when Manchester tries to be sympathetic about MacArthur's military brilliance, you get the feeling MacArthur absolutely failed upwards
Secretary of Defense Rock@sodrock.bsky.social · 2d agoWe welcome any and all MacArthur slander
confidence bands around nonlinear curves can be misleading/tricky too because you run into pointwise versus curvewise uncertainty mjskay.github.io/ggdist/refer...
Curvewise point and interval summaries for tidy data frames of draws from distributions — curve_interval
Translates draws from distributions in a grouped data frame into a set of point and interval summaries using a curve boxplot-inspired approach.
mjskay.github.io
Fascinating and clear paper by @corymccartan.com and @melodyyhuang.bsky.social, greatly enhancing our understanding of how Bayesian Additive Regression Trees (BART) works and why it is so effective. A must-read for my fellow BART enthusiasts. #statssky #causalinference
ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · 3d agoarXiv📈🤖 Seeing the Forest for the Trees: The Gaussian Process Limit of BART By McCartan, Huang
When I took a management class 20-ish years ago, this book was one of the prof's suggestions as an "anti-leadership/anti-management book". A lot of the lessons apply beyond the military realm.
Foreign Policy@foreignpolicy.com · 2d agoA 50-year old book reveals the tendencies that undermine military competence—a century ago and today. Read our analysis for free:
This discussion is weird. I've seen a lot of "things are lower quality/cheaper nowadays than in the past" and...maybe? No one seems to be adducing anything besides vibes (to Isaiah's point)
Isaiah Bishop@isaiahbishop.bsky.social · 3d agoChances are you are comparing very durable antiques for a lot of stuff, and not the regular shit. Like you know how rare chairs were 200 years ago?
Good advice, even for us non-professors
Clément Canonne@ccanonne.github.io · 4d ago"We're all worried," as what it means to do research (in my field, Theoretical CS) seems to be shifting, and shifting fast. What to do? Senior researchers must lead by example, knowing that not everything will pan out. What I'm suggesting below may not work everywhere, but here's my own advice: 1/
The problem is that "immortal time bias" is not an entertaining enough name. Maybe "vampire bias" or "zombie bias" is better
Julia M. Rohrer@dingdingpeng.the100.ci · 3d agoReading up on immortal time bias and wow, I don't know whether there's any other bias that has a name that makes it so hard to explain what it actually is?
I do not even remotely understand what is going on anymore.
Isaiah Bishop@isaiahbishop.bsky.social · 4d agoPeople are saying this person is crazy so lets wait a bit
Hey #rstats: Is there a way in Positron to view variables outside the Global Environment? In Rstudio I'll frequently change the focus environment in the environment pane to explore the contents of loaded packages, but I can't figure out how to do that in Positron. 🤔 @ivelasq3.bsky.social ?
Blades of Steel for the original NES
Sophia_Nyx 🏳️⚧️@sophianyx.bsky.social · 5d agoName a videogame that deserves a comeback.
Yes. I have not worked in industrial research, only at universities and govt agencies. But I have good friends in biotech, pharma, and Silicon Valley, and the stories I hear from them have done as much as the theory of public goods to convince me of the vital importance of NIH and NSF.
Joshua G. Schraiber@jgschraiber.bsky.social · 5d agoNothing convinced me more of the importance of publicly funded basic research more than my time in industry.
Would appreciate some #rstats feedback on a potential `tinyplot` change: github.com/grantmcdermo... Should we default to suppressing the (redundant) inner axes for faceted plots? We already do this for themes without a plot frame (box). But I'm thinking of enabling for framed plots too. #dataviz
There's an interesting tension here. If humans have name for it, then it's a social construct. But some social constructs are us trying to make sense of the world around us (e.g., ADHD, carnivores, sleep), and others are wholly constructed by us (e.g., the law, the market, the nation-state).
Stephen Wild@stephenjwild.bsky.social · 5d agoSomething can be both a brute fact and a social construct. The distance between points A and B? Brute fact. Unit of measurement to describe the distance (cm, inches, rods, parsecs)? Social construction. Thank you for your attention to this matter!
Really a job title, filled by whoever is currently putting lines through things, until they fail and are replaced by the next guy. Stephen Wild, the James Bond of Bayesians.
This is all I can think of (laudatory)
Jess Calarco@jessicacalarco.com · 5d agoFor example: When I teach qualitative methods, I ask my students which of these foods are sandwiches, and there's always lots of disagreement. So, we talk about how they can explain their criteria and try to persuade others to agree with them. But they can't objectively prove they're right.
And the reason people socially construct brute facts is that those constructions allow us to coordinate action. If we don't have units of measure, we can't easily give each other directions. If we don't have labels for symptom clusters like ADHD, we can't easily develop treatments or policies.
Stephen Wild@stephenjwild.bsky.social · 5d agoSomething can be both a brute fact and a social construct. The distance between points A and B? Brute fact. Unit of measurement to describe the distance (cm, inches, rods, parsecs)? Social construction. Thank you for your attention to this matter!
In economics, the philosophical concept that gets egregiously misused out of the seminar room is "rational."
Olúfẹ́mi O. Táíwò@olufemiotaiwo.bsky.social · 5d agothere are a lot of philosophical concepts that get egregiously misused out of the seminar room (and in the seminar room, to be fair). but in terms of sheer pernicious misuse "social construct" has gotta be top 5 somewhere and it's not number 5
Something can be both a brute fact and a social construct. The distance between points A and B? Brute fact. Unit of measurement to describe the distance (cm, inches, rods, parsecs)? Social construction. Thank you for your attention to this matter!
Olúfẹ́mi O. Táíwò@olufemiotaiwo.bsky.social · 5d agoyour apartment, the sport of basketball, and the Republic of Korea are all social constructs. mere recognition of this fact does not actually make the question of whether you are standing in your own apartment, playing basketball, or a Korean national up in the air until settled by op ed editors
A bunch of you are talking about a thing I am knowledgeable about, and you are all wrong. This has been a practice vague post. Or was it? Yes, it was. But really, was it? bsky.app/profile/laur...
lauren@lauren.rotatingsandwiches.com · 6d agoalso don't go harassing people who feel differently about this. the vaguepost is a more honorable tool than the weaponized quote dunk
The obvious solution is to use linear or Poisson regression and skip odds ratios entirely.
Julia M. Rohrer@dingdingpeng.the100.ci · 6d agoNew blog post! Odds ratios are...a bit weird. Here I explain two problems (interpretation, non-collapsibility) and provide some simple recommendations for improved reporting. www.the100.ci/2026/07/30/r...