Lightnews — Scholar-powered news

Richard Hooper @richardhooper.bsky.social · 5d

This is the best I can do: www.bbc.co.uk/programmes/p...

BBC Two - Parade's End - Christopher Tietjens

Played by Benedict Cumberbatch

www.bbc.co.uk

1

Richard Hooper @richardhooper.bsky.social · 5d

Trial statisticians (this may be a well-worn/tedious question)—do you have views/experience of integrating randomised allocation sequences in Qualtrics? Is the built-in randomisation module in Qualtrics fit for purpose? Are there ways of integrating, say, stratified block randomisation?

1

Richard Hooper @richardhooper.bsky.social · 7d

But modelling quantifies the uncertainty in the data. Here (even if AI is doing what people imagine it does) the training data are plonked in place of the real world, like an unconvincing background painting being audio-described.

1

Richard Hooper @richardhooper.bsky.social · 7d

This is apropos of nothing, but if the plural of quiz is quizzes, why is the plural of bus buses?

Richard Hooper @richardhooper.bsky.social · 7d

Slightly open-mouthed on reading this pre-print (thanks @statsepi.bsky.social):

"If achievable, this ...[using LLMs] would enable researchers to predict the responses of vulnerable and/or hard-to-reach populations of participants without the need for time- and resource-intensive research studies."

Darren Dahly @statsepi.bsky.social · 7d

Science is grounded in observation. Measurement is a tool for observation. Measurements should be evaluated for validity and reliability/uncertainty. Scientists who use measurements without understanding their properties are not really scientists at all.

Jamie Cummins @jamiecummins.bsky.social · 20d

Can large language models stand in for human participants?
Many social scientists seem to think so, and are already using "silicon samples" in research.

One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.

THREAD 🧵

1 3 5

Richard Hooper @richardhooper.bsky.social · 7d

You can find details of other webinars in the series here: www.nihr.ac.uk/news/nihr-la...

www.nihr.ac.uk

1

Richard Hooper @richardhooper.bsky.social · 7d

I'm doing a webinar in two weeks (15 Oct 1-2pm BST), part of NIHR Research Support Service's Methods Academy series, on the age-old question: should I do a stepped wedge trial?

Free registration, details here: tinyurl.com/z6cjvce3

@nihr-rss.bsky.social @wiphcem.bsky.social @qmul-wiph.bsky.social

1 15 14

Richard Hooper @richardhooper.bsky.social · 8d

I really enjoyed this talk from @richarddriley.bsky.social today, particularly the football quotes! Looking forward to seeing this available on the NIHR YouTube channel soon. @nihr-rss.bsky.social

Richard Riley (R²) @richarddriley.bsky.social · 8d

Giving a talk today which anyone can watch online: 1pm - 2pm
"The Importance of Sample Size for the Development and Evaluation of Clinical Prediction Models"
- intended for a broad audience (anyone doing or using prediction model research)
imperial-ac-uk.zoom.us/webinar/regi...

Welcome! You are invited to join a webinar: Research Methods Academy: Seminar - The Importance of Sample Size for the Development and Evaluation of Clinical Prediction Models. After registering, you w...

Led by Professor Richard Riley Join this Online Seminar on clinical prediction models estimate individual risk for outcomes from disease. Many studies lack rigorous design; this session will focus o...

imperial-ac-uk.zoom.us

1 3 2

Reposted by Richard Hooper

Jess Butler @jessbutler.bsky.social · 8d

To be way too earnest for on here, I genuinely think nerds being careful is political activism these days and it cheers me up

1 2

Reposted by Richard Hooper

Richard Riley (R²) @richarddriley.bsky.social · 8d

Giving a talk today which anyone can watch online: 1pm - 2pm
"The Importance of Sample Size for the Development and Evaluation of Clinical Prediction Models"
- intended for a broad audience (anyone doing or using prediction model research)
imperial-ac-uk.zoom.us/webinar/regi...

Welcome! You are invited to join a webinar: Research Methods Academy: Seminar - The Importance of Sample Size for the Development and Evaluation of Clinical Prediction Models. After registering, you w...

Led by Professor Richard Riley Join this Online Seminar on clinical prediction models estimate individual risk for outcomes from disease. Many studies lack rigorous design; this session will focus o...

imperial-ac-uk.zoom.us

7 13

Reposted by Richard Hooper

Stepped Wedgehog @steppedwedgehog.bsky.social · 13d

📢 Registration now open for the 11th annual meeting on Current Developments in Cluster Randomised Trials & Stepped Wedge Designs on 18th & 19th November 2025 in Birmingham UK.

shop.bham.ac.uk/conferences-...

2 1

Richard Hooper @richardhooper.bsky.social · 18d

Our wild rocket went to seed in the heat of the summer and it's still putting on a beautiful show.

Reposted by Richard Hooper

Hetan Shah @hetanshah.bsky.social · 26d

Pet parrots which typically live alone (whilst those in the wild live in large flocks) were given the technology to call each other. They would use it for up to three hours a day, and developed favourite friends 💔
on.ft.com/3K05vhS

When a parrot wanted to connect with a distant friend, a touchscreen showed a selection of other birds available online. The parrots learned to activate the screen, designed specially for them, by touching it gently with their tongues rather than pecking aggressively with their beaks.
"We had 26 birds involved," said Hirskyj-Douglas. "They would use the system up to three hours a day, with each call lasting up to five minutes." The interactions ranged from preening and playing with toys to loud vocal exchanges.
"When we went through the data, we found that most of these parrots had favourite friends," said Hirskyj-Douglas.

130 2.1K 6.7K

Richard Hooper @richardhooper.bsky.social · 26d

Why do I hope to see this earlier? It speaks to the design of the trial, and the scientific basis for that; it's one of the first things you think about in the development of the trial, and therefore something I would expect to see early in the "story" of the trial.

Richard Hooper @richardhooper.bsky.social · 26d

Usually bundled with some other, inconvenient "statistical considerations"

Reposted by Richard Hooper

Stepped Wedgehog @steppedwedgehog.bsky.social · 26d

Save the date: Fifth International Conference on Stepped Wedge Design 2026 12th-14th May (online)

www.york.ac.uk/healthscienc...

1 1

Richard Hooper @richardhooper.bsky.social · 26d

When I come across a new template for a clinical trial protocol it's fun to see how far I have to read before I get to the sample size justification. Section 16.1 Sample Size Calculation!—about par for the course.

2 2

Richard Hooper @richardhooper.bsky.social · 27d

Having an opportunity to lead their own projects; also, finding out more about career paths that were new to them, such as trial management and data management

1

Richard Hooper @richardhooper.bsky.social · Sep 8

Thank you to our four amazing undergrad interns, Maryam, Anoushri, Sandhiya & Sky for joining us in @wiphcem.bsky.social this summer to experience health research in action. Thanks also to @nihr.bsky.social for funding through their UIP and the RSS Hub at Imperial & Partners.
@qmul-wiph.bsky.social

4 7

Reposted by Richard Hooper

Journal of Clinical Epidemiology @jclinepi.bsky.social · Sep 4

Our updated @jclinepi.bsky.social starters pack!

go.bsky.app/DNtLZWo

5 5

Richard Hooper @richardhooper.bsky.social · Aug 28

NIHR Research Support Service @nihr-rss.bsky.social has announced a webinar series on research methods (innovation/best practice), aimed particularly at those applying for UK funding for health research who want to bridge the methods implementation gap, but open to all. More topics/dates to come.

NIHR Research Support Service (RSS) @nihr-rss.bsky.social · Aug 28

We are excited to launch our new Research Methods Academy: Seminar Series!
Hear from leading experts on innovative and best practice methods to strengthen your research proposal. Find out more and register: www.nihr.ac.uk/news/nihr-la...

Richard Hooper @richardhooper.bsky.social · Aug 28

I would dearly like to cut out the "when I first get up" check: why is so much email traffic from my UK colleagues sent during the night? Yes, I've seen the signatures: "I choose to work these funny hours but I don't expect you to", yada yada -- you're still stressing me out, people!

1 2

Reposted by Richard Hooper

Peter Ellis @freerangestats.info · Aug 26

I deleted "data scientist" from my LinkedIn profile and replaced it with "statistician". Data scientist somehow feels very 2010s, and whatever it "really" means, it is certain to cause more confusion about what I do and my skillset than does the older word.

4 1 32

Richard Hooper @richardhooper.bsky.social · Aug 22

When you walk into the gents' toilet at our workplace, the light only comes on in the neighbouring, disabled toilet, and when you walk in there the light comes on in the gents'. Somewhere an electrician is probably still chuckling to themselves.

1

Richard Hooper @richardhooper.bsky.social · Aug 22

I feel like your cat would be slumped too, if it weren't for the "recovery suit".

1