Spending some time with family in the Research Triangle area and looking to connect with folks interested in research methods, causal inferences, and AI ethics. Ping if this is you! I'm working on mass screenings for low-prevalence problems, science + statistics reform, uncertainty, and feedbacks.

Looks like a targeted screening programme to me. Can't find a recommendation from the National Screening Committee either way on this one. Hope it is being looked at because of the risk of harm and waste.

Rejoice! Papa Bear has been found again. His reappearance spurs me to think harder about endpoints, what really matters -- and how the liberty-security sequential-feedback continuum may extend to medical case studies. But where fore-art thou, adequate data? wildetruth.substack.com/p/bear-is-ba...

but that’s fine. It served a useful function at the time in getting people to think about potential informational value in P-values - it just was too crude an approach, so stimulated development of other methods. Can we have a “z-curve for dummies” explainer to improve take-up?

Arguably the first and most important use of language: getting others to do things. (Conveying factual information, on the other hand: not universally acquired and unstable without training.)

Pediatrician: Does your six month old say "baba"? Me: My baby says "hi," "hey," "Mama," "I go," and ("Are you gonna go get your brother?") "I'm gonna"!

I got a free copy of *The Precipice: Existential Risk and the Future of Humanity* from 80,000 Hours. Glancing at the table of contents, it fails to recognize perspectival limits including bias and error in the evidence on which it's based. But, on p. 6 so far, it's pretty good. Ruin it for me.

So much this. Except I started five blog posts in the midst of a mild autoimmune flare, childcare going away for a while, and grandparents flying in from the States for a summer visit. Now I am not sure which if any of us will survive the next two weeks.

Day 5 with my mother. Please send help. Chocolate, flame-throwers -- every bit helps.

Causal inference iceberg! What's missing?

41 MEPs call upon the European Commission to initiate an independent inquiry into large-scale unlawful transfers of personal data between @frontex.bsky.social and @europol.europa.eu following our investigation for El Pais, Le Monde, Solomon, and Netzpolitik.

Remember the librarian problem from Tversky & Kahneman? Neither does Joanne Briggs — daughter of fraudster scientist Michael Briggs. So I'm reading re-relevant, vintage Brian Deer on failed HIV vaccine trials instead of her new memoir. wildetruth.substack.com/p/birth-cont...

New @eckles.bsky.social et al. piece out in Biometrika and straight onto my summer reading list: arxiv.org/pdf/2004.09458 Maybe I'll finally get a grip on the relationship between continuity- and randomization-based identification strategies in RDD.

Your world may end in fire; theirs may end in ice. But mine will end in butter and salt. Won't that be nice? (with deference to Robert Frost)

Bad proxies are bad. Friends don’t let friends confuse easy metrics with meaningful outcomes. But how do we do better? From breast cancer screening to mass surveillance for child sexual abuse material, it's not entirely obvious what endpoints should be. wildetruth.substack.com/p/bad-proxie...

Oh yes new jazz / romantic piano fix, oh yes Emahoy Tsegué-Maryam Guèbrou. www.youtube.com/watch?v=nKU7...

Continual moral panic over companies (gasp) selling formula is a mommy-shaming scam that hurts babies. Moms commonly can't produce enough milk to feed their babies, buy formula to do the job, and then are told they're not doing what's "best." Going hungry isn't better. See verawil.de/breastfeeding

wildetruth.substack.com/p/you-say-to... Frustrated with how iterated screening models often throw away info, I fell into a Bayesian search rabbit hole so you don't have to! Spoiler: it means different things to different people, and multilevel modeling may make more sense.

Astral Codex Ten goes deep on the missing heritability question and recent findings on educational attainment www.astralcodexten.com/p/missing-he...

A lost bear -- found. A New Yorker article -- flawed. An endpoints rainbow for vaccines -- sketched. And a question about how to extend that to security and other rule-abidingness domains. (Or: Wait, what was the question?) wildetruth.substack.com/p/the-bear-t...

The bear is lost. The bias is pervasive. Maybe I can finally manage to accept defeat as nobly as my four year old. wildetruth.substack.com/p/we-didnt-s...

Illuminating the process of producing rigorous science by selecting on the dependent variable

the omnisolution:

Relevant blog post by @ruben.the100.ci: www.the100.ci/2020/06/24/m...

Two salvaged Covid vaccination essay drafts, one last science writing post before I really try to do the tech skilling up pivot I should've done years ago. wildetruth.substack.com/p/notes-from...

Indeed another excellent and timely piece to unpack the discourse behind it. “The lesson from abroad is clear: lasting change requires transforming the conditions that make parenthood feel untenable, not just offering token incentives.”

Read what happened in English with partner @technologyreview.com www.technologyreview.com/2025/06/11/1...

I’ve spent a lot of time investigating shitty AI cynically used to pursue vulnerable people in welfare systems. When I talk about this work, I invariably get asked: is it possible to make these systems fair?

Currently popular Chez Wilde: - *Press Here* www.amazon.de/-/en/Press-H... & - *The Book with No Pictures* www.amazon.de/-/en/Book-No... Both are children's meta-books -- books based on a premise of functioning in a different way from how books normally function. Is this a new phenomenon?

IJ Good on varieties of Bayesians (many) and a Bayes/non-Bayes compromise (practical), h/t Sander Greenland. fitelson.org/probability/...

Recent health + methods readings, ft. the usual methods mistakes & spin: - Covid vax in pregnancy may increase birth defect risks. - Abortion, antidepressants, & RSV vax in pregnancy may increase preterm birth risk. - Underpowered trials get spun as conclusive. wildetruth.substack.com/p/links-0611

Really interesting, very in-depth on how open science and replication efforts were funded and gradually grew more influential. Great example of what thoughtful philanthropy can accomplish. goodscience.substack.com/p/metascienc...

In which I explain why (and not everyone has to agree): statsepi.substack.com/p/out-of-bal...

3/4 plants agree having a baby in January was survivable!

I suppose it makes sense if you believe testing has no harms (it does), overdiagnosis has no harms (it does), and overtreatment has no harms (it does). The public just hear "getting screened saved my life!", and it's difficult to have a detailed conversation with a GP when there's no time.

Supporting Scientific Data | A Guide for Researchers I'm writing up the guidance and recommendations that I have put together after almost a decade of supporting research data management. Putting this out there because why not and to put pressure on myself to finish up the remaining chapters.

I wish there was a library of resources on research integrity methods and resources... Oh! Look.... 👍👏

Friends, I have news. It is very good news. You'll just have to read it, I'm afraid. jamesclaims.substack.com/p/introducin...

Post-publication peer review in action:

Science-integrity project will root out bad medical papers ‘and tell everyone’ Thrilled to announce this new $900,000 project headed by @jamesheathers.bsky.social

What is all this about shared medical decision-making? I know the consumer model has its problems, but don't most people just decide for themselves what treatment they need and source it? wildetruth.substack.com/p/my-doctor-...

The four-year-old saw my everything profile photo (for messaging, social media, blog...) and asked "Who's that?" Jeez, it's from January. Apparently four months with two kids changes a person!

This week, the CDC stopped recommending Covid vaccination for healthy kids and pregnant women. Experts like Peter C. Gøtzsche and Vinay Prasad have long said that makes sense. But Paul Offit calls it part of "RFK Jr.'s War on Children." Hyperpolarized much? wildetruth.substack.com/p/when-exper...

A recent paper claims ~45% better performance for mammography when machine learning selects who to screen, ft. fewer invasive cancers and less overdiagnosis. But it inflates the AUC gain, neglects causality, and raises questions about informed consent. wildetruth.substack.com/p/miracle-or...

Most frame abortion as life and death. Both sides prioritize different lives. But what if the epistemic structure of the problem is that of a mass screening for a low-prevalence problem? That would mean we can estimate type I and type II errors. wildetruth.substack.com/p/abortion-a...

Just out: ALGORITHMIC HARM. www.amazon.com/Algorithmic-...