Carrboro Coachella, y'all.

John Rees (Rhys)@jrees.net · 4w ago

It’s Tomato 🍅 Day at the Carrboro Farmers Market

Look, I have already named my alternate-career bakery (the Pastriarchy) but someone really needs to open one called Intention To Treat.

That feeling of excitedly finding a paper that solves your current methods headache, only to realize you found the paper last year and fully memory-holed it.

I can't help but think that the "Think of AI as an eager grad student" people just... don't like working with grad students very much.

Next week: ISPE webinar on avoiding immortal person time bias in pregnancy medication studies (link below to register, registration is free and open to non-ISPE members). us06web.zoom.us/webinar/regi...

Welcome! You are invited to join a webinar: Avoiding Immortal Time in Studies of Medication Use During Pregnancy. After registering, you will receive a confirmation email about joining the webinar.

Immortal time persists in studies of medication use during pregnancy conducted with real-world data. Even among experienced pharmacoepidemiologists, the challenge often lies not in defining immortal t...

us06web.zoom.us

It's not 1536, but Anne Boleyn has taken over the news. Recently a team of computer scientists, electrical engineers and a historical novelist think they have discovered the “true face” of Anne (& her mother) by applying AI-based computational facial recognition to Holbein sketches. The catch? /1

Sonja Drimmer@sonjadrimmer.bsky.social · 3mo ago

They’re at it again. None of this is valuable historical research. It’s about exploiting historical artifacts and historians to develop computer vision further and to make a soft cultural argument for surveillance tech. This nonsense has been going on since c2010.

The quotes on this one are a gift.

s. e. smith@sesmith.lol · 4mo ago

Do not cite the deep magic to me, witch, I was there when it was written

Important work!

ashlee, mph@avs.bsky.social · 4mo ago

we’re back up and running! we’re doing a two-stage approach with a screener and then i will send a one-time link to participants. if you have had a miscarriage since march 2024 and otherwise meet requirements as listed on the flyer, please consider participating in my survey (go.uic.edu/leavestudy)

Peter over here posting about his favorite pizza, which is definitely a sandwich

Peter Tennant@pwgtennant.bsky.social · 4mo ago

I tend to get dogpiled every time I say this but: I like ham and pineapple pizza.

Teaching sibling comparison designs on Friday, which means I get to bust out this slide:

Oh no oh no oh no

Julia M. Rohrer@dingdingpeng.the100.ci · 5mo ago

Risky click

The official journal of the Canadian Paediatric Society has just acknowledged that more than 100 of its case reports are fabricated. Incredible reporting from @retractionwatch.com retractionwatch.com/2026/03/03/c...

Not all heroes etc etc

Andrew Heiss@andrew.heiss.phd · 6mo ago

research haikus, you say?? www.andrewheiss.com/research/

But I want a hippopotamus for Christmas. Only a hippopotamus will do.

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

This is a hell of a year to be staring down the tenure clock, having yet to secure 'R01 or equivalent' funding. I've submitted 10 proposals this year: 6 as PI, 2 foundations. It's weird to feel both anxious about this, and also guilty for feeling anxious about this amid all the *gestures wildly*

Bill Hanage@billhanage.bsky.social · 8mo ago

Essential reading if you care about the health of scientific inquiry in the US www.nytimes.com/interactive/...

I went to undergrad with a guy who thought eating was a waste of time, so he blended and drank all his meals. I can’t remember his name, but the tuna fish sandwich shakes will be branded into my brain until the day I die.

Andrew Heiss@andrew.heiss.phd · 8mo ago

Discovered new way to quickly consume leftover pies: blend half gallon of ice cream + some milk + an entire pie in a kitchenaid. Between the 7 of us, a whole pie-in-shake-form disappeared in like 90 seconds

I'm hiring a researcher to work on several ongoing funded projects at the University of Oslo! Details below: www.jobbnorge.no/en/available...

Researcher in PharmacoEpidemiology and Biostatistics (289818) | University of Oslo

Job title: Researcher in PharmacoEpidemiology and Biostatistics (289818), Employer: University of Oslo, Deadline: Tuesday, December 2, 2025

jobbnorge.no

Potential outcomes tabs in Outlook! DAGs in your Google query! Marginal structural models to help you plan your next vacation!

Darren Dahly@statsepi.bsky.social · 9mo ago

The advances we've made in statistics, experimental study design, and causal inference over the past century are remarkably useful for understanding our world. But there is never been a push to make people use them like we are seeing with generative AI. Perhaps take a moment to consider why.

I have a weekly meeting with a USian who works in Norway and this is the first time our 10-minute politics-related therapy session at the top of the hour has been hopeful, so.

TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10/NSF-... 🧵

The official home of the Python Programming Language

python.org

Ah yes, let's try the "have you tried turning it off and on again" approach to fixing US healthcare.

J. Emory Parker 🏳️‍🌈@jaspar.bsky.social · 10mo ago

UPDATE: Hours after saying they were pausing all payments, Medicare says they’re NOT pausing all payments. More ➡️ www.statnews.com/2025/10/15/c...

Weekly Ruport: bad news cycle means more emotional support pillows

It’s hard to know whether talking about these things is helpful or just gives them more oxygen, but the current acetaminophen fire is already roaring so here we go. Most importantly: ACOG affirms their position that acetaminophen is safe in pregnancy bsky.app/profile/acog...

ACOG@acog.org · 11mo ago

ACOG reaffirms that acetaminophen is safe for managing pain and fever during pregnancy. No reputable studies support suggestions like those in HHS’s recent announcement linking acetaminophen use in pregnancy to autism; in fact, high-quality studies show no such risk. https://bit.ly/47Wxc59

I'm hoping for advice from NIH funded researchers re the updated OA policy (grants.nih.gov/grants/guide...). Specifically: I have students funded on an F31 or T32 whose papers were submitted this spring and accepted after 7/1. Per policy, those papers must be OA from pub date... (1/3)

Props to the CDC staff who presented the science at yesterday’s #ACIP meeting. Key points relating to #CovidVaccination in #pregnancy… 1. No increased risk of any problems in pregnancy, at birth, or for infants 1/ www.cdc.gov/acip/downloa...

I don't know how many of my ISPE friends (and enemies...?) are on here, but (a) the ISPE Peer Mentoring program is currently taking applications and (b) it's GREAT, you should probably sign up. Link: www.pharmacoepi.org/get-involved...

The thing about middle age is that one of your hips always lies and the other always tells the truth.

I went to a talk last year and remember thinking: did he just say we should use AI to create study samples? Not for simulation studies, but to answer an applied question? I must have misheard, that’s clearly bonkers. Anyway, here we are.

Verasight@verasight.io · 12mo ago

Our new analysis, in collab with @gelliottmorris.com, finds major errors when replacing humans with synthetic samples. Over the next 5 days, we’ll share some of the findings. https://report.verasight.io/synthetic-sampling/?utm_source=bluesky&utm_medium=organic&utm_campaign=launch&utm_content=cover

1. The philosophy of science sometimes gets an unearned reputation as a purely academic exercise that offers little by way of concrete tools for advancing research. This is wrong. And today, as we grapple with how AI is changing the nature of scientific activity, it's desperately wrong.