What If We Reversed Bayes to Test How Robust Our Findings Are to Skepticism?

In this week's issue of Causal Python Weekly:

- Lu Qian's new blog post on Reverse Bayes and evidence

1/

#CausalSky #EconSky #EpiSky #MLSky #StatSky

ICHPS 2027 Invited Proposal Please click the link to complete this form.

(2) If you have ideas related to the role of #statsky, #AI, #realworlddata, or #digitalhealth in policy, we’d love to see them! Please share with your networks. #hpsr #hspr @drjwolfson.bsky.social @jaredhuling.bsky.social @mboudreaux.bsky.social @amstatnews.bsky.social @enar-ibs.bsky.social

Is the effect heterogeneity in your data real?

Testing effect homogeneity and confounding with mixed data

In their new working paper, Ana Paula Armendariz and Martin Huber propose a framework for testing the...

1/

#CausalSky #EconSky #EpiSky #StatSky #MLSky

You Need an Analysis. Which Statistical Test to Use?

Hey there, why not ask Claude?

With LLMs, the bottleneck is no longer the speed of access to information or the speed of content generation, but the effort needed to verify the outputs.

1/

#StatSky #EpiSky #BioStats #CausalSky

Digital twins in healthcare need 100x as much derision as these stupid LLM survey respondents. We can barely identify average causal effects in patients we literally intervene on, but you're gonna simulate a whole set of individual treatment courses? It's Neopets for doctors. #statsky #episky

Dropbox Reallocates $25M After a Causal Study

In a recent paper, Varun Chivukula et al describe a series of large-scale blackout experiments across marketing channels that inspired the company to reallocate $25M of their marketing budget to more effective channels.

1/

#CausalSky #EconSky #StatSky

Dropbox Reallocated $25M of Their Marketing Budget After a Causal Study.

In this week's issue of Causal Python Weekly:

- Alberto D. Horner reviews a new End-to-End Estimation for Counterfactual Fairness by Yuchen Ma et al.

1/

#CausalSky #EconSky #EpiSky #StatSky #MLSky

5 Causal Inference Ideas for the Age of Vibes

I'm working on a talk for an event this spring, and I want to ask you for your thoughts.

I collected these 5 ideas in response to various queries I got from people across industries and...

1/

#CausalSky #StatSky #EpiSky #EconSky #MLSky

Any good resources for visualising the posterior interval of, e.g., a propensity for treatment model? If you get 10k draws per patient and have thousands of patients, it could be useful to show individual prediction intervals in a clean format. Pinging @andrew.heiss.phd
#rstats #bayes #statsky

I fall in love with @gt-package.bsky.social every time I use it.

I remember back in my early days when I’d manually make tables by looking at values in SAS outputs. Glad that is behind me. #rstats #statsky

Is the Effect Heterogeneity in Your Data Real?

In today's issue of Causal Python Weekly:

- A new paper by Ana Paula Armendariz and Martin Huber on testing for heterogeneity and confounding in high-dimensional mixed data (observational + experimental)

1/

#CausalSky #StatSky #EpiSky #EconSky

DHS Seeks Access to Massive Employment, Salary and Family Database Legally Restricted to Use in Child Support Cases The Trump administration’s immigration enforcers are requesting unfettered access to the government’s “most powerful people-finder system,” which contains sensitive information on children and victims...

I attended multiple sessions at #FCSM2026 about low response rates to survey. I am 100% confident that ICE raids cause fear among respondents. Federal data are not meant to exploit people's basic needs. Thanks, @csmarcum.bsky.social for the discussion.
#statsky #srms #defenddata #defendresearch

Hivemind: I want to know the increase in survival time from repeated Tx visits over uneven times. Pts can be followed up repeatedly, and Tx can be given over multiple visits. I'm after time to each visit and time-to-outcome. What would you do? Weibull frailty model? This is in brms. #statsky #rstats

Very Handy!

The latest version of CausalPy introduces a new piecewise interrupted time series (ITS) class.

CausalPy is a popular Python library for modeling quasi-experimental data.

The new version comes with a brand new...

1/

#CausalSky #EconSky #EpiSky #MLSky #StatSky

Having built a lot of (tabular data) clinical prediction models over the past few years, it's astonishing how close you get to the AUC of most ML models just by whacking a spline on your important continuous variables. rcs() go brrrr
#rstats #statsky

I have to self-teach most of the stats and econometrics I apply in my work. SEO has made finding good resources incredibly frustrating, and if LLMs can help me quickly learn a concept, show me how it works in R, and find a good book/article on the topic, it's an unambiguous win. #statsky #econsky

National statistics are in crisis around the world — and the impacts will be severe Some researchers are sounding the alarm over the official data sets that track crucial aspects of life in the United States, Argentina, the United Kingdom and India.

I've been at #FCSM2026 today because federal data matter. @clairemkbowen.bsky.social is cited here, and there's no paywall needed to read this! #defenddata #defendresearch #statsky #AIsky

Mamiya-Sekor C 500mm f5.6 for Mamiya 645 The Mamiya-Sekor C 500mm f5.6 is a very big telephoto lens for the Mamiya 645 medium format camera. In excellent condition, with light cosmetic wear- most notably some light scuffing on the lens hood-...

Just in: "it's not every day you see a 500mm lens for Mamiya 645" they say, but today is a day, and this is a 500mm lens for a Mamiya 645, and if we extrapolate from that one data point we can safely say with 100% certainty that you will see a 500mm lens for a Mamiya 645 every day. #statsky #science

Truncated count models should come with a bright, flashing warning light (and maybe some sort of horn) alerting users to the complexities of coefficient interpretation.

Or maybe raw coefficient interpretation should just be outlawed? #rstats #statsky

What can the Clinical Frailty Scale tell us about planning for death? By Dr Robin Blythe and Prof. Adrian Barnett. Blog post for palliAGED and CareSearch, the Palliative Care network.

Contributed a blog to CareSearch/palliAGED on our external validation of the #clinicalfrailtyscale. The CFS is very widely used but we showed limited prognostic value. Better to think of it as a cue for discussing ageing and death with older pts than as a marker to assign treatment
#statsky #ageing

"Just Plot the Data"

Which of the two datasets is normally distributed?

Plotting the data might seem like the most assumption-free way to examine its properties.

But is that really the case?

1/

#StatSky #Biostatistics #EconSky #EpiSky

International Conference on Health Policy Statistics The American Statistical Association is the world’s largest community of statisticians, the “Big Tent for Statistics.

Reminder to #hpss #hpsr #hspr #episky #statsky #AIsky #RWDsky that the #ICHPS2027 proposal deadline is approaching. Submit a workshop or session idea. @csmarcum.bsky.social and @cynthiarudin.bsky.social are our keynote speakers!

What a Week!

We literally had too many topics to fit into one newsletter this week.

Here's what we picked:

- Alberto D. Horner reviews the brand new book by Quentin Gallea, PhD

- David Rohde on why policies are stochastic in reinforcement learning

1/

#CausalSky #StatSky #EconSky #EpiSky #MLSky

Interactive resources With the power of OJS and Quarto, I’ve created a few interactive websites to illustrate trickier statistical concepts when teaching. Check them out (and adapt and copy as much as you want!) With links to three different websites (accessible at the main link in the post)

Finally got around to adding fancy links to my different interactive teaching websites for showing things like p-hacking, p-value interpretations, and (still-in-draft-form) DAGs at www.andrewheiss.com/teaching/ #rstats #QuartoPub #statsky

Modern Science Relies on an Idea That Disproving Is Easier Than Proving.

The idea was proposed by Karl Popper nearly a century ago.

We call it "falsificationism"

A million white swans don't prove all swans are white.

Neither does a trillion.

But...

1/

#StatSky #EpiSky #Biostatistics #EconSky

A medical journal says the case reports it has published for 25 years are, in fact, fiction A Canadian journal has issued corrections on 138 case reports it published over the last 25 years to add a disclaimer: The cases described are fictional. Paediatrics & Child Health, the journal…

🚨 OMG! No words!

#academicsky #idsky #statsky

Last Friday, We Launched "Intro to Biostatistics" with Justin Bélair

Two ideas inspired us to build it:

1/

#StatSky #Biostatistics #EpiSky

Can someone point me to published comment on the pros/cons of standardised risk of bias tools vs question specific tools?

#EpiSky #StatSky

logitr: Fast Estimation of Multinomial and Mixed Logit Models with Preference Space and Willingness-to-Pay Space Utility Parameterizations by John Paul Helveston <p>This paper introduces the logitr R package for fast maximum likelihood estimation of multinomial logit and mixed logit models with unobserved heterogeneity across individuals, which is modeled by a...

Doing a discrete choice experiment for the first time and {logitr} by @jhelvy.bsky.social has saved the day and my sanity. Takes data frames, easy to understand, and crucially, incredibly fast for multinomial logit. We are so lucky that people do this for science and not profit
#rstats #statsky

2027 International Conference on Health Policy Statistics (ICHPS) For over 25 years, the International Conference on Health Policy Statistics (ICHPS), organized by the Health Policy Statistics Section of the ASA, has played a vital role in the dissemination process ...

#ICHPS2027 is now accepting proposals for invited workshops and sessions. Come join us in Annapolis. The deadline to submit your proposal is March 31, 2026. The theme is Evidence-based Policy in Action: Statistics, AI, and Digital Health in Practice. #hpss #hspr #hpsr #statsky #AIsky #RWD