Doesn't look good for those over 25. What is going to happen to them? And why so suddenly?

One week away! Don't miss the next Methods Series presentation with Alex Ocampo.

The median age in China has rapidly caught up with the United Kingdom. Line chart of median age for China and the United Kingdom from 1950 to 2025, with the vertical axis in years from 0 to 40 and the horizontal axis showing years 1950 to 2025. A line labeled United Kingdom stays around mid-30s in 1950, dips slightly to about 33 by the mid-1970s, then gradually rises to about 40 by 2025. A line labeled China starts around 22 in 1950, falls to about 18 to 19 in the mid-1960s and 1970s, then climbs steadily to meet the UK at about 40 in 2025. Annotated note: in the mid-1960s China’s median age was just under half that of the UK; another note states that today the median age in both countries is 40 years. Data source: UN, World Population Prospects (2024). License: CC BY.

The median age in China has rapidly caught up with the United Kingdom—

In 1965, the median age in the United Kingdom was almost twice that of China. Half of the people in the UK were younger than 34 years, and half were older. In China, this midpoint was just 18 years.

nzz-artikel: "IT-Branche Zürich lockt Frauen: Pink Die IT-Branche will mehr Frauen für sich gewinnen – mit viel Pink und Rosa und «<einer bildlichen, einfachen Sprache>> Eine Studie, unterstützt vom Kanton Zürich, <deckt die wahren Empfindungen von Mädchen und Frauen auf». Zeno Geisseler 26.05.2024, 05.35 Uhr 3 min" darunter ein bildausschnitt einer mutmaßlichen frau mit pinkem hemd (man sieht nur einen teil des torsos, arme und hände), die in einem büro sitzt und einen laptop bedient

least sexist industriebranche

The Geordi LaForge meme demonstrating hesitation for a conventional “causal” direct graphical model, showing only observed variables and unobserved confounders, and enthusiasm for a directed graphical model that represents a full joint probabilistic model including all observed variables and all latent/unobserved variables.

Meanwhile, Spain's massive investment in renewables is paying dividends now: with prices for Spanish industry and consumers low and stable compared with other European economies.

www.ft.com/content/ac77...

May I use Proton?

... with birds visiting and gliding effortlessly through the air

Post a pic you took, no context, to bring some zen to the feed.

Confidence intervals with maximal average power We propose a frequentist testing procedure that maintains a defined coverage and is optimal in the sense that it gives maximal power to detect deviations from a null hypothesis when the alternative to...

Bookmarked. Need to read carefully.

FYI, The figure is very similar to arxiv.org/abs/1905.03981, in which we use repeated sampling from hypothesis, and assess the use of priors for frequentist confidence intervals.

DAG representing the causal structure of a standard difference-in-differences design with two locations and two time periods—units in one location in the post-period receive treatment. $L$ = group or location indicator (treated vs. untreated location); $T$ = time indicator (pre vs. post period); $U$ = unobserved time-invariant confounders (e.g., GDP per capita, general health status, public health infrastructure). $X \leftarrow T \rightarrow Y$ represents a common time trend affecting both locations equally. The causal effect of $X$ on $Y$ is identified by conditioning on $\{L, T\}$, which corresponds to using location and time indicator variables in a regression like `y ~ location * period`.

DAG representing the causal structure of a standard difference-in-differences design, but with explicit pre- and post-treatment outcomes. $L$ = group or location indicator (treated vs. untreated location); $T_\text{post}$ = post-period measurement (indicator that the observation occurs after the intervention); $X_\text{post}$ = treatment (which only occurs for treated locations in the post period); $Y_\text{pre}$ and $Y_\text{post}$ = outcome measured before and after the intervention. $U$ = unobserved time-invariant confounders (e.g., GDP per capita, general health status, public health infrastructure). $Y_\text{pre} \rightarrow Y_\text{post}$ represents outcome persistence (e.g. autocorrelation or slow-moving changes); $X_\text{post} \leftarrow T_\text{post} \rightarrow Y_\text{post}$ represents a common time trend affecting both locations equally. The causal effect of $X_\text{post}$ on $Y_\text{post}$ is identified by conditioning on $\{L, T_\text{post}\}$, which corresponds to using location and time indicator variables in a regression like `y ~ location * period`.

spending my sunday evening once again attempting to draw a DAG for diff-in-diff

📣The 2026 Symposium of #CausalInference in the #HealthSciences takes place on March 18, 2026 in Fribourg. Theme: AI & machine learning in causal inference for health sciences
🎤Keynotes: Elsa Gautrain, Aurélien Sallin, Jonas Peters, Jana Mareckova
🔗https://projects.unifr.ch/pophealthlab/?page_id=1561

Some Common Dose–Exposure–Response Estimands and Conditions for Their Causal Identifiability Exposure–response analyses are central to dose selection in drug development. The estimand framework, formalized in ICH E9(R1) regulatory guidance, provides a structured approach to define scientific...

We apply the estimand framework to dose–exposure–response analyses. ... strategy to improve exposure–response analyses for dose selection, particularly when the relevant evidence includes data from multiple studies.

#estimand #exposure-response #dose-response #causal

doi.org/10.1002/psp4...

Confidence Regions for Multiple Outcomes, Effect Modifiers, and Other Multiple Comparisons In epidemiology, some have argued that multiple comparison corrections are not necessary as there is rarely interest in the universal null hypothesis. From a parameter estimation perspective, epidemio...

For years I had trouble following some of the discussion about confidence bands, but at ACIC this year @noahgreifer.bsky.social pointed me to a helpful paper

So you don't have to be as perplexed as I once was, we have a new pre-print introducing the key ideas
arxiv.org/abs/2510.07076

adrftools: Estimating, Visualizing, and Testing Average Dose-Response Functions Facilitates estimating, visualizing, and testing average dose-response functions (ADRFs) for characterizing the causal effect of a continuous (i.e., non-discrete) treatment or exposure. Includes suppo...

I'm so excited to announce the first release of my newest #Rstats package, {adrftools}! This package facilitates estimation, visualization, and testing for the causal effect of a continuous (i.e., non-discrete) treatment.

🧵 1/10

#statssky #episky #causalinference

Thanks for these posts. I would have answered that live would be boring if you can only analyze pre-registered compliant data.

Your answer makes much more sense.

👏

As I said, this is the important work that needs to be done. Mark's paper is overly simplistic, arguing we can't judge deviations at all. Deviations are in practice not remotely as bad as he wants. If he had collected actual data,che would have falsified his own claims.

Here an example of using DAG/SWIG to assess a situation with an underlying process that is continuous in time, and interventions and observations that are discrete.

bsky.app/profile/chri...

Ok ... if the data is a sample from a larger population, then it may be convenient to describe the relation between the two via a distribution from which the data is sampled.

Ok ... the functional form by which the independent variables affect the dependent variable matters. When you use linear models as a tool, you may have to transform the independent variables.

This is a fairly technical but highly relevant paper on how we can model complex systems at various levels of detail without losing causal content. Think gas: instead of tracking every molecule, we can focus on big-picture properties like temperature and pressure. www.auai.org/uai2017/proc...

Question that intrigued me a lot. My current view:

Arrows can represent the process that is continuous in time. Edges are values of the process at selected time points of particular interest.

We have to be careful. Sometimes, a data summary may nonetheless answer a different question, and this different question could be of interest.

Don't do anything since there might be a bias might be counterproductive!

Why it matters
1️⃣ Clear estimand definition – the target of inference is stated up front, removing ambiguity.
2️⃣ Transparent causal assumptions – DAGs & SWIGs show which confounding paths are blocked.
3️⃣ Step‑by‑step DER derivation – fits into the standard dose‑exposure‑response workflow.

2/6

What are some of the best DAGs you seen that depict time-varying confounding?

Could be from a 'real' worked example, or generic.

#EpiSky

Or as joint outcomes? With the mediator helping to explain variability of the outcome?

We propose a method for estimating long-term treatment effects with many short-term proxy outcomes: a central challenge when experimenting on digital platforms. We formalize this challenge as a latent variable problem where observed proxies are noisy measures of a low-dimensional set of unobserved surrogates that mediate treatment effects. Through theoretical analysis and simulations, we demonstrate that regularized regression methods substantially outperform naive proxy selection. We show in particular that the bias of Ridge regression decreases as more proxies are added, with closed-form expressions for the bias-variance tradeoff. We illustrate our method with an empirical application to the California GAIN experiment.

arXiv📈🤖
Long-Term Causal Inference with Many Noisy Proxies
By Lal, Imbens, Hull

Uruguay’s Renewable Charge: A Small Nation, A Big Lesson For The World Uruguay built a power grid that runs 99% on renewables—at half the cost of fossil fuels. Here’s how its bold energy overhaul became a global model.

“Uruguay did what most nations still call impossible: it built a power grid that runs almost entirely on renewables—at half the cost of fossil fuels. The physicist who led that transformation says the same playbook could work anywhere—if governments have the courage to change the rules.”

X/Twitter's rough full volume is around 500 total million posts every day, or 182 (and a half) billion posts per year.

By contracts, we found 11.2 million research posts in all of 2025 on there.

In other words, 0.000006% of Twitter appears to be sharing research. Basically zero.

This gives financial institutions leeway to acquire hard-to-sell, higher-yielding long-term assets and finance them with cheaper short-term liabilities, thus increasing profits through a risky “mismatched book”.