Matti Vuorre | Lightnews

Matti Vuorre @matti.vuorre.com · 2h

And that's a fine conclusion too (if we can be confident in it) and should directly inform the next experiment (probably need way more participants then.)

Matti Vuorre @matti.vuorre.com · 3h

Thanks, interesting paper. I'm missing a conclusion in the example write-up. Is it that the hypotheses (JND & prior SD) are so uncertain that there cannot be a conclusion? I find this generally the stickiest point against interval tests.

Matti Vuorre @matti.vuorre.com · 20h

4 of 17 member institutions are in the Netherlands 🇳🇱

Society for the Improvement of Psychological Science (SIPS) @improvingpsych.org · 1d

PsyArXiv is pleased to welcome new member institutions!
✨Eindhoven University of Technology
✨New York University
✨Tilburg University
✨University College London

Member institutions are integral to helping PsyArXiv deliver preprint posting and reading to its users at no charge.

Society for the Improvement of Psychological Science (SIPS) @improvingpsych.org · 1d

PsyArXiv is pleased to welcome new member institutions!
✨Eindhoven University of Technology
✨New York University
✨Tilburg University
✨University College London

Member institutions are integral to helping PsyArXiv deliver preprint posting and reading to its users at no charge.

Matti Vuorre @matti.vuorre.com · 1d

I am not averse to the S value, and I like how it converts the question to one of quantifying information. It would be lovely to see some experiments evaluating whether it could deliver (so comparing interpretations & decisions against p values etc.)

Matti Vuorre @matti.vuorre.com · 1d

Looks great, I'll read tmrw. I am not sure what the panda thing is though! avi why not use a more modern video codec

Matti Vuorre @matti.vuorre.com · 1d

😢

Matti Vuorre @matti.vuorre.com · 1d

Who said stats is hard!

Matti Vuorre @matti.vuorre.com · 1d

I just encountered this for the first time last week: I could view but not download a paper. wtf

Lukas Röseler @aufdroeseler.bsky.social · 1d

Now this is a new one: My university's license of SAGE journals allows me to view but not to download the research articles. According to ChatGPT read-access costs 100K€ p.a. If you want to make 100% sure that interested readers can access your article, publish it in a #diamondopenaccess journal.

Matti Vuorre @matti.vuorre.com · 1d

So if I flip it I can then say null is equally probable as x *tails* in a row 🥴

Matti Vuorre @matti.vuorre.com · 1d

If the null is true these results (or more extreme) have the same surprisal value as x heads in a row? Am I reading this right?

Ben Bolker @bbolker.bsky.social · 1d

A colleague is looking for an open-source/online (pref. peer-reviewed) reference for properties of probability distributions. Any ideas? (NIST gives *very* basic properties e.g. www.itl.nist.gov/div898/handb... but I think they're looking for something more complete ...)

I’m working on a manuscript that mentions a fair number of standard distributions and their properties, and I’ve been wondering if there’s an equally standard reference. For context for my question, while Wikipedia often has what I need if I want to appeal to some function and its properties, the Digital Library of Mathematical Functions (https://dlmf.nist.gov/) provides a peer-reviewed standard alternative. Is there someplace similar where I could point for properties of a Gumbel, Dirichlet-multinomial distribution, etc.

Lukas Röseler @aufdroeseler.bsky.social · 1d

Now this is a new one: My university's license of SAGE journals allows me to view but not to download the research articles. According to ChatGPT read-access costs 100K€ p.a. If you want to make 100% sure that interested readers can access your article, publish it in a #diamondopenaccess journal.

Robert (Bob) Kubinec @rmkubinec.bsky.social · 6d

🚨 New draft 🚨

We built an LLM-enabled system to measure greenwashing scores in 1 million worldwide Facebook ads.

We found vast networks of Facebook pages sharing pro-fossil fuel messages & show that ads are targeted at left-leaning areas with fossil fuel investments.

Link: doi.org/10.48550/arX...

links of pages with connections to fossil fuel companies and similar types of climate change messaging

Matti Vuorre @matti.vuorre.com · 7d

The MSFT angry bird 🫡

Matti Vuorre @matti.vuorre.com · 7d

It's always sunny in philadelphia.

Matti Vuorre @matti.vuorre.com · 7d

Microdosing hot peppers! We need more science on this asap.

two police officers shake hands in front of a sign that says fx

ALT: two police officers shake hands in front of a sign that says fx

media.tenor.com

Matti Vuorre @matti.vuorre.com · 7d

For anyone who hasn't seen it yet, season 17 was very good. It brought back a lot of the elements from earlier seasons that made the show what it is (though not all the minor characters I'd have hoped for). I also appreciate the 8 episodes per season format--quality over quantity.

Matti Vuorre @matti.vuorre.com · 7d

Incredible dataviz.

Dannagal Goldthwaite Young @dannagal.bsky.social · 7d

Time for another episode of “bad data viz”

This one truly has me here thinking “Am I a Dolt?” What is this even saying??? Doesn’t WSJ have amazing talented staff skilled in data visualization? Like.. what is this?

Incomprehensible graph/picture about Rx use among kids

Dannagal Goldthwaite Young @dannagal.bsky.social · 7d

Time for another episode of “bad data viz”

This one truly has me here thinking “Am I a Dolt?” What is this even saying??? Doesn’t WSJ have amazing talented staff skilled in data visualization? Like.. what is this?

Bastian Jaeger @bxjaeger.bsky.social · 7d

It's easy to produce spurious findings:
A meaningless score based on irrelevant evaluations (“My relationship has very good Saturn”) was moderately related to common relationship measures (satisfaction, commitment) & predicted those measures 3 weeks later

journals.sagepub.com/doi/10.1177/...

Matti Vuorre @matti.vuorre.com · 8d

So 'working on files on my computer' <-> syncing with a cloud system (github et al.) -> archiving the files and directories (zenodo et al). It's not rocket science!

Matti Vuorre @matti.vuorre.com · 8d

I agree. This might have something to do with supporting "modular" research projects / publishing, but at the very least the UI does not help. I support using something called a 'file system' where projects can have 'directories' and 'files', and those can be easily shared & collaborated on :)

Matti Vuorre @matti.vuorre.com · 8d

This post highlights how far behind the OSF is in usability and speed. The ResearchBox UI does seem clean, but why not use Zenodo directly (RB archives to Zenodo anyway.)

Data Colada @datacolada.bsky.social · 8d

[130] Researchbox: Even easier to use than before. More transparently permanent.
datacolada.org/130

[130] ResearchBox: Even Easier to Use and More Transparently Permanent than Before - Data Colada

Over the past 10 years or so, posting data, code, and materials for published papers has gone from eccentric to mundane. There are a few platforms that enable sharing research files, including Researc...

datacolada.org

Matti Vuorre @matti.vuorre.com · 8d

Thanks!

Matti Vuorre @matti.vuorre.com · 8d

Time to stop doing non-open reviews or at least complain to the journal? That sucks.

Matti Vuorre @matti.vuorre.com · 8d

How strong is strong? Posteriors for 50% n=2,4,8,16 plz 🙏

Matti Vuorre @matti.vuorre.com · 8d

Yeah makes sense. Also probably 'sensitive' behaviors affected differently / for longer time. What if I just never tell the participants 🥸

Matti Vuorre @matti.vuorre.com · 8d

This episode reminded me of 2 things:

1. We know how & why the bootstrap works, but the fact that it does is just very cool.
2. Bayes. All that strapping (& which bias adjustment to choose) just goes away as unnecessary.

Quantitude the Podcast @quantitude.bsky.social · 17d

Join us tomorrow as we repeat ourselves.
Join us as we repeat ourselves tomorrow.
Tomorrow, we repeat ourselves. Join us!

Data Colada @datacolada.bsky.social · 8d

[130] Researchbox: Even easier to use than before. More transparently permanent.
datacolada.org/130

[130] ResearchBox: Even Easier to Use and More Transparently Permanent than Before - Data Colada

Over the past 10 years or so, posting data, code, and materials for published papers has gone from eccentric to mundane. There are a few platforms that enable sharing research files, including Researc...

datacolada.org

Add to Home Screen

Light up
your news

Add to Home Screen

Light upyour news

Sign in to Lightnews

Sign up to start reading

Connect Bluesky

Connect with Bluesky

Light up
your news