Georgia Tomova
@georgiatomova.bsky.social
Research Fellow at @clscohorts.bsky.social epidemiology, causal inference, methods, meta-science Subscribe to the Causal Inference Interest Group: https://ciig-seminar.github.io
I'm very excited to hear this critical examination of causal triangulation! Triangulation is something that I have always thought is a 'good idea', but which I've been somewhat skeptical about, due to a lack of formalism.
Georgia Tomova@georgiatomova.bsky.social · yesterdaySuper excited for our next CIIG seminar on causal triangulation with Keling Wang 💃 See the abstract and register here: ucl.zoom.us/meeting/regi... See more about the CIIG and subscribe: ciig-seminar.github.io
Super excited for our next CIIG seminar on causal triangulation with Keling Wang 💃 See the abstract and register here: ucl.zoom.us/meeting/regi... See more about the CIIG and subscribe: ciig-seminar.github.io
I'm just a statistician Standing in front of an analysis plan Wondering why the outcome was defined so it could only occur in one exposure group
They hounded Aaron Swartz to death for downloading papers from jstor but society is now supposed to just accept mass theft because this thing can write that TPS report *for* you. en.wikipedia.org/wiki/Aaron_S...
Aaron Swartz - Wikipedia
en.wikipedia.org
John Scalzi@scalzi.com · 4d agoThis is your occasional reminder that generative "AI" in particular could have been developed on public domain materials and/or licensed and compensated works but the major players intentionally, methodically and systemically chose to steal labor from working creative folks instead
Sophisticated statistics can produce extremely precise nonsense when the design fails to identify the causal mechanism.
"We'll probably have flying cars in 2026" 2026: Triple-checking online stock images to make sure people don't have 7 fingers.
"Is an ultra-processed chocolate cake materially more harmful than a home-made one?" I enjoyed contributing to this response to The Lancet series on ultra-processed foods, which outlines some key limitations in the evidence that 'ultra-processing' is specifically harmful to health.
Public health policy should be grounded in established evidence that acknowledges uncertainty: a response to The Lancet series on ultra-processed foods - BMC Medicine
Background The recent three-part Lancet series on ultra-processed foods (UPFs) argued that the displacement of traditional diets by UPFs is a major driver of poor diet quality and diet-related chronic...
link.springer.com
Small-N assessment: AI makes everyone more productive output-wise, but I, as a senior person with prior knowledge benefit the most, while my ("junior") students struggle with actual learning. This paper proves this with a larg(er)-N causal design: www.nber.org/papers/w35720
Does AI Assistance Enhance or Erode Expertise? Evidence from a Three-Month Field Experiment in Patent Drafting
Founded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers, an...
nber.org
We're at @socsocmed.bsky.social's conference all this week, for the first time! Our cohort data are featuring across a wide range of talks and posters. If you're there, come say hi to our friendly team at our stand. @georgiatomova.bsky.social @spilleron.bsky.social
Causal inference is the art of knowing exactly why your estimate isn't causal
The venue for the RSS conference this year is the Bournemouth International Centre (BIC). Nice, I always prefer BIC to AIC.
Does the new shingles vaccine (Shingrix) protect against cardiovascular disease (CVD)? Yes, according to a new 'natural experiment' study in @natmed.nature.com. But the truth is sadly far less convincing! Pull up a chair and let's examine another peer review failure in a Nature journal. 1/10 🧵
In nutrition research, the effect you estimate from a model is determined by which other dietary variables you adjust for. If you adjust for total energy intake, the effect will switch from an addition to a substitution. Which substitution will be determined by which other variables are adjusted.
People who use the term "covariable" - why? Do you use it synonymously with "covariate"? If so, why not say "covariate"?
📢 Join our experts for an overview of the challenges – and opportunities – involved in making causal inferences using @clscohorts.bsky.social cohort data. Book now ➡️ buff.ly/gbWzhwh
How many nutrition meta-analyses pool estimates for incomparable research questions, due to different energy adjustment strategies? We looked at an illustrative sample. The answer was all of them! Delighted to share our new meta-scientific review, led by @nataliaortega.bsky.social 👇
Meta-analyses of dietary exposures must consider energy adjustment: recommendations from a meta-scientific review
Abstract. In observational studies of dietary exposures, the energy adjustment strategy directly alters the effect being estimated. Adjusting for total ene
academic.oup.com
📣 Job alert: We are looking for a PhD Candidate in causal inference & health data science🔎 📍 Berlin (Responsible Health Data Analysis Team, BIH-QUEST) 📅 Starting: Oct 1, 2026 (3 years) 💼 100% E13 position 🔗 Read more and apply (deadline Aug 17): karriere.charite.de/en/job-vacan... Please share!
PhD Candidate (all genders) BIH QUEST Center (Team “Responsible Health Data Analysis”)
full-time | 17.08.2026 | BIH Berlin Mitte
karriere.charite.de
There is a major problem with how applied health (equity) researchers think about covariate control. I often read research papers where the paper is trying to answer an important question....But the adjustment set doesn't make sense. #episky 1/3
I think one of the most underrated skills of a good PhD supervisor is actively teaching you the hidden curriculum in academia.
New paper out now 🥳 When psychologists discuss generalisability, they often refer to vague notions of representativeness. We provide an accessible intro to the total survey error framework as a tool to reason about this more rigorously. w @taymalsalti.bsky.social @ruben.the100.ci >
Academia is like constantly playing the final level of those games where you run a restaurant and you have 50 burgers to make and you don't manage to serve anyone before their patience depletes.
“Descriptive” is a property of the research question, not the perceived quality of the results. I often see people state a study limitation and then say “therefore the results are descriptive”. You can’t do an analysis first and then use the results to define what type of study you wanted.
