DON'T MISS this seminar next week. It will *not* be recorded! 👇

Psychologists doing clustering be like: #statsmeme

Tell me you don't understand Bland-Altman limits of agreement without telling me. Authors: "93.5% (100/107) of pairs fell within 95% limits of agreement"

New blog post! Let's say an effect you are interested in varies between countries and you want to explain some of that cross-cultural variation. What could possibly go wrong? www.the100.ci/2026/05/20/f...

"Standard research practice in medical research" = "Is this research misconduct"?

As ifffff

Strange world where a Bulgarian Olivia Rodrigo lookalike singing nonsense saves us from literally weeks of discourse.

Are any biomedical research studies based on data that are not from the real world?

You: please act normal you need these people to like you Me: ok! People: hi Me: Did you know that John Snow did not draw his famous cholera map during the 1854 Soho outbreak itself, but in the months afterwards to help explain his water transmission theory to colleagues and the wider world.

You can now watch Jeremy's excellent talk on the term "association" and what it's even supposed to be referring to on YouTube. Must watch if you're doing associational research! www.youtube.com/watch?v=IKdC...

Why we should rethink causal mediation, and what to do instead? Come to hear the answer from Vanessa Didelez at the next CIIG seminar! The seminar will be hybrid. If you are in London, come join us in person at UCL! Otherwise, you can join on Zoom as usual. Registration links in comment below.

The recording of this record-breaking (in attendee numbers) seminar is now available 🎉

If anyone else is also missing out on our epic seminars, here is where you can subscribe and also see previous recordings 😉 ciig-seminar.github.io

New collider bias example just dropped

Important point that comes up over and over again: Don't conflate "strong questions" (does X affect Y) with "strong answers" (X absolutely surely certainly affects Y). You can absolutely start with a causal question and then conclude that the data permits any strong conclusions.

Jeremy right now

Starting soon! :)

A slide from this talk I'm giving on Monday. The conclusion will be a set of questions that one must be able to answer to ask associational research questions. The ball is in your court association people!

I'm a mid-career researcher insofar as my research career is mid

Starting @bjsociology.bsky.social conference with 3 papers on social mobility and an old Hungarian joke: "Communism is the longest and most painful transition from capitalism to capitalism"

Do you ever claim to be interested in the "association" between variables? This talk may be for you! @jeremylabrecque.bsky.social will explain when and why such associational language is not appropriate and why causal language should be used instead. Register here: ucl.zoom.us/meeting/regi...

naming your biology plagiarism machine after the most famous case of biology plagiarism is a very funny bit I must admit

I think policies that support editors to desk reject more papers could free up capacity Dichotomise continuous variable - reject Stepwise variable selection - reject Table 2 fallacy - reject Etc If widely applied, this could drive up standards

I’ve been posting and talking about mode effects a lot recently, but if you still don’t know what mode effects are and/or how to deal with them in your analyses, now is your chance to learn! Register to attend our FREE online event on 30-Apr👇🏻 @rjsilverwood.bsky.social @clscohorts.bsky.social

Science is a cumulative and collaborative process that requires human connection and interaction. I find it interesting that the 'lone genius' idea still gets romanticised in academia.

Recording now available if you've missed it:

TODAY at 3pm UK time! Register here: ucl.zoom.us/meeting/regi...

Replicated finding ≠ true* Unreplicated finding ≠ false** *because internal validity biases can and do replicate **because external validity refers to a specific population - the size/presence of an effect in a particular study is dependent on effect modifiers in the original sample/population

To gain false confidence in your results? The awkward fact is that structural biases replicate. The HRT and CVD studies of the 1980s all replicated the same horrifically biased result that HRT protects against CVD risk. This is a key argument for triangulation.

Late to the conversation, but I'm kind of unbothered by this lack of replication. If the study is not internally valid to begin with, then being able to replicate it only means replicating systematic biases. And something tells me this is likely to be a bigger problem in the literature.

economists be like "our worst quality is that we're very good at this."

Basically: academia is so hostile to a healthy personal life that the only way to have a romantic relationship and/or parent children is to bond with another academic?

It’s really quite shocking Wellcome didn’t have internal policy/feedback systems that prevent burning considerable resources on pseudo-science…

Our preprint is now out! 📄 We looked at whether five flexible IPTW / propensity score weighting approaches can be used for count exposures, and whether they can be combined with multiple imputation. We also illustrate the methods using BCS70 data. Link: doi.org/10.48550/arXiv.2603.23726

I love the internet

I do not want to use AI to write or think. Can someone train it to handle annoying tasks like formatting journal submissions instead of trying to get us all to use it for things that take critical thought and expert knowledge?

oil prices are about to make polyester a noble fiber

A very warm welcome to Bluesky to @helena-koerber.bsky.social! 🥳 Helena is a very talented early-career researcher with interests in research methods, surveys, and social attitudes. Please follow her and suggest the best accounts/feeds for her to follow!😊

Why do so many PhD students to this day get taught to "find the gap in the literature"? It has nothing to do with understanding what questions *need* answering.

Alternately the conclusions could be phrased "We spent many hours and tax payers money on a project to which we have at best made a very minor scientific contribution that better research is needed." Note the word "better" rather than "more".

Risky click

if you get this reference to early-'20s twitter, log off, touch grass, and burn your copy of The Book of Why

Bonus points: who are the compliers to a genetic instrument?

Another week in which I ask myself who are the compliers in a complier average causal effect and why are they our target population

Meta-analyses in nutrition may pool estimates for completely incomparable effects without realising it, just because of whether/how energy intake is adjusted. @nataliaortega.bsky.social will explain why, share results from a meta-scientific review & offer recommendations for ways forward 👇

📢 Researchers at @ucl.ac.uk propose new approach to improve quality of population data research by looking at data linked from the Millennium Cohort Study to the UK National Pupil Database (NPD). Read the full paper on the @ijpds.bsky.social journal website 👉 buff.ly/HJLkFkg

If you missed this seminar, you can: 📹 watch the recording here: youtu.be/n5HHg4RJQns?... 📖 read the (brand new!) paper here: doi.org/10.23889/ijp... cc @rjsilverwood.bsky.social @clscohorts.bsky.social @sriucl.bsky.social

This was very neat - paper here ijpds.org/article/view...