Lightnews — Scholar-powered news

LightNews

Dan Lewer

@danlewer.bsky.social

290 followers 230 following 270 posts

Consultant in Public Health at Bradford Institute for Health Research. Interested in quantitative research methods and mental health

Posts Media Videos Starter Packs

Dan Lewer @danlewer.bsky.social · 3h

Thank you!!

Julia M. Rohrer @dingdingpeng.the100.ci · 4h

New blog post!

Let's say you have two measures meant to capture the same confounder. They're highly correlated. Can you still proceed with your regression analysis?

(I admit, the title is a bit of a spoiler)

www.the100.ci/2025/10/13/i...

If you have two measures of the same confounder, you can just include both of them in your regression model

Sometimes, researchers worry about multicollinearity in situations where it’s actually a non-issue. Here’s one such scenario. Imagine a situation where you are interested in the effect of X on Y (X...

Dan Lewer @danlewer.bsky.social · 10h

Chi-square tested univariate associations

Dan Lewer @danlewer.bsky.social · 1d

In honour of spooky month, share a 4 word horror story that only someone in your profession would understand.

The p-value was 0.051

Daniel Johnson @danjohnsonhymns.bsky.social · 1d

“Well, a comment, actually”

Marjorie James Keenan @unenthusiast.com · 1d

In honour of spooky month, share a 4 word horror story that only someone in your profession would understand.

rm -rf ~/

Dan Lewer @danlewer.bsky.social · 3d

For me the #1 bugbear is "there's no way we could afford an RCT so we did some kind of non randomised study and analysed it wrong, which appeared statistically more powerful than a trial of the same size"

Dan Lewer @danlewer.bsky.social · 3d

I think the difficult part is "avoidable". Some studies you think "that could have been done differently", others you think "there was no way of answering that question given your data/resources"

Dan Lewer @danlewer.bsky.social · 3d

That's a very basic persona. I'm maintaining an "anti-tidyverse" persona (tidyverse is a specific package within R)

Dan Lewer @danlewer.bsky.social · 3d

Google came up with something useful by moving the data around -maybe some kind of advanced bootstrapping

Dan Lewer @danlewer.bsky.social · 3d

It did include confidence intervals so that's a step forward

Dan Lewer @danlewer.bsky.social · 3d

I asked chatgpt to analyse the graph

Dan Lewer @danlewer.bsky.social · 3d

omg 100% I was also being sarcastic about that insane regression ... But would it be fun to be a data journalist? Their plots look so much nicer than ours

Dan Lewer @danlewer.bsky.social · 3d

VIX Index and FT Macro mood seem like robust variables tho ... and they do report the gold standard for causal analysis: % of variance explained

Dan Lewer @danlewer.bsky.social · 3d

The more I use R, the more I realise they did a pretty good job of deciding what you probably need in the basic package #RStats

Dan Lewer @danlewer.bsky.social · 3d

To be really cynical about it, I'd argue that the people most enthused by LLMs in quantitative research are those who never really liked stats and methodology

Dan Lewer @danlewer.bsky.social · 4d

Top tip for journal editors and funding bodies who review evaluations of public health interventions: rather than send it to peer reviewers, just ask "did you realise your intervention is clustered?"

Dan Lewer @danlewer.bsky.social · 5d

I think historically bubbles have continued for a while even when everyone knows it's a bubble. So people carry on investing even though they know it's pointless, as they believe someone else will be even more foolish. But how does that explain Google tripling capex and abandoning net zero

Dan Lewer @danlewer.bsky.social · 5d

Is that an experienced professor in the background behind them?

Dan Lewer @danlewer.bsky.social · 6d

Genuinely odd comparison!

Our World in Data @ourworldindata.org · 6d

Does the news reflect what we die from?

The image presents a comparison of the leading causes of death in the United States for 2023 and the media coverage these causes receive from three news outlets: The New York Times, The Washington Post, and Fox News.

In the footer, it notes the data sources, indicating that the information is based on media mentions from Media Cloud (2025) and death data from the US CDC (2025) and the Global Terrorism Index, with a clarification that values are normalized to sum to 100%.

Dan Lewer @danlewer.bsky.social · 6d

It should be happy to help if you tell it you're an acceptable member of society, eg. A researcher

Dan Lewer @danlewer.bsky.social · 6d

You could take it even more obscure and do Windows98%CI

Dan Lewer @danlewer.bsky.social · 6d

Overall I agree, but in the UK the answer would be that vaping has not been approached as harm reduction, but rather an exciting new product that is marketed at children. This is obviously a policy failure rather than an inherent problem with the technology (an own goal)

Dan Lewer @danlewer.bsky.social · 6d

Don't listen when older people tell you to enjoy your youth. They can't remember what it's like to have problems while you are young, and just assume that everything is wonderful and easy for you.

Reposted by Dan Lewer

Rafe Meager (they/them) @economeager.bsky.social · 6d

jumpscare on instagram @dandekadt.bsky.social

some instagram account has posted a screenshot of dan's post that says "My wife and I never talk baby talk to our kid. We enforce academic seminar norms and jargon. My child isnt even 2 yet and can interrupt presenters on the first slide. Adults at family gatherings are shocked when my child asks a questino that turns out to be more of a comment." it has THIRTY SIX THOUSAND SEVEN HUNDRED AND FIVE LIKES

Dan Lewer @danlewer.bsky.social · 7d

Oh I see. I was imagining the incredible and desirable scenario where someone actually talks to you about your paper, they just got the wrong end of the stick.

Dan Lewer @danlewer.bsky.social · 7d

That's the dream!! Someone read your paper??

Dan Lewer @danlewer.bsky.social · 10d

Do you think the pilot episode "I have some data - can you help me analyse it" would lead to a commission?