Lightnews — Scholar-powered news

Reposted by Quentin André

Tim Kaiser @tkaiser.science · 23h

Cool and normal

Quentin André @quentinandre.bsky.social · 6d

(The Cochran-Mantel-Haenszel approach is NOT an interaction test: It tests whether an association between two variable exists after controlling for a stratifying variable).

2

Quentin André @quentinandre.bsky.social · 6d

If, as a reviewer, you see an unusual statistical test being reported in a manuscript, it never hurts to double-check whether the test does what the authors claim.

Who knows, it might help a journal not publish inaccurate results.

2 2

Reposted by Quentin André

Hakeem Jefferson @hakeemjefferson.bsky.social · 26d

I have many thoughts about Charlie Kirk—and perhaps even more about the white elites, including some on the left, who insist we can’t hold multiple realities at once. We can. And we must.

A brief 🧵

1 130 440

Reposted by Quentin André

Jack Lynch @jacklynch000.bsky.social · Aug 17

I'll admit, I was skeptical when they said Gemini was just like a bunch of PhDs. But I gotta admit they nailed it.

Google's Gemini AI tells a Redditor it's 'cautiously optimistic' about fixing a coding bug, fails repeatedly, calls itself an embarrassment to 'all possible and impossible universes' before repeating 'I am a disgrace' 86 times in succession

71 1.7K 7.3K

Reposted by Quentin André

Marco Zocca @ocramz.bsky.social · Aug 5

holy shit lmao

1 1 5

Quentin André @quentinandre.bsky.social · Jul 9

I host it on my personal website: quentinandre.net.

A few example of "open science" posts:
quentinandre.net/post/circlin...
quentinandre.net/post/is-powe...
quentinandre.net/post/making-...
quentinandre.net/post/large-p...

Quentin André

Quentin André<br>Assistant Professor of Marketing

quentinandre.net

2

Quentin André @quentinandre.bsky.social · Jul 9

I don't know what counts as "open science", but I've written a bit about things like interpreting effect sizes and p-values, ensuring reproducibility...

1

Quentin André @quentinandre.bsky.social · Jul 8

1

Quentin André @quentinandre.bsky.social · Jul 8

Like with Ego Depletion and Stereotype Threat, Construal Level Theory has, following failures to replicate, been "motte-and-bailey"-ed to more modest claims. In the meantime, many scientists have wasted time and effort on it, rather than collecting interesting facts about human behavior.

1

Quentin André @quentinandre.bsky.social · Jul 8

Take Construal Level Theory: The idea that different instantiations of distance (physical, social, temporal, perceptual...) are all psychologically related (Liberman and Trope 2008).

This Big Beautiful Idea was formulated on little data... but hundreds of studies quickly "confirmed" it!

2/3

1 3

Quentin André @quentinandre.bsky.social · Jul 8

Very interesting post, which largely aligns with my views on theory.

I see another danger with Big Beautiful Theories: In a world in which researchers' degrees of freedom are insufficiently constrained, they become self-fulfilling.

1/3

Michael Inzlicht @minzlicht.bsky.social · Jul 8

I just became president of a scientific society and my first act was to disagree with one of my heroes...Arie Kruglanski gave a keynote arguing social psychology needs MORE Big Beautiful Theories. I think that's exactly backwards. Here's why...

Hasty Theories

Why Social Psychology Needs to Stop Rushing to Explain

open.substack.com

1 2

Quentin André @quentinandre.bsky.social · Jul 8

More from HBS' Amended Complaint. They claim to have evidence that Gino backdated a fabricated file to exonerate herself.

5 10

Quentin André @quentinandre.bsky.social · Jul 8

Key claims from HBS

3 3

Quentin André @quentinandre.bsky.social · Jul 8

Do you remember Francesca Gino's claim on her website that HBS analyzed the "wrong data file" in their investigation, and that a "real file" proved her innocence?

HBS is now claiming that the "real file" was fabricated by Gino... and thus that Gino's claim was defamatory.

6 16 84

Quentin André @quentinandre.bsky.social · Jul 7

Looks like the "Tariff" button is back on the resolute desk.

1

Quentin André @quentinandre.bsky.social · Jul 7

Super interesting thread. A stark reminder that any metric of "variance explained" is basically uninterpretable without contextualizing (i) the magnitude and variability of the treatment and (ii) the "baseline" variability of the outcome.

Julia M. Rohrer @dingdingpeng.the100.ci · Jul 7

"If we could jointly analyze extended Korean families, we would likely find that the influence of the shared environment – which side of the border you ended up on – is not only real, even for something like height, but possibly trumps genetics."

Ruben C. Arslan @ruben.the100.ci · Jul 7

Another contribution in the heritability debate. Makes a point I agree with: many of the important environmental causes vary by place and time and are marginalized in most h2 studies.
open.substack.com/pub/easthunt...

1 2

Quentin André @quentinandre.bsky.social · Jul 2

I'll believe the Gen AI hype when a LLM can write a 10 page report without using the "This isn't just X: It's Y" structure 8673 times.

This isn't just annoying, it's aggravating.

2

Reposted by Quentin André

Nick Brown @steamtraen.eu · Jun 25

Of course academics should not sue others for pointing out their faulty (*cough*) work.

But this is a bit like saying that mafiosi should not threaten witnesses. Doing that follows from being a mafioso, just as suing people who point out your fraud is part of the kind of person who commits fraud.

1 2 7

Quentin André @quentinandre.bsky.social · Jun 12

If you haven't seen this... southpark.cc.com/video-clips/...

Flip the Pyramid - South Park | South Park Studios US

In the midst of the gluten crisis, Cartman calls the USDA with a crucial piece of information.

southpark.cc.com

Quentin André @quentinandre.bsky.social · Jun 8

Francesca Gino doesn't do fraud.

She just has a lot of fraud happening to her.

(shamelessly stolen from a friend).

JAMES HEATHERS @jamesheathers.bsky.social · Jun 8

“My promise is to work as hard as I can to present the material as fairly and clearly as I can to establish both the innocence of Francesca Gino and the injustice of the process that has led to this result.”

Now, that’s two entirely separate things.

substack.com/home/post/p-...

In defense of Francesca Gino

An argument to be made here, in words and a podcast

substack.com

2

Quentin André @quentinandre.bsky.social · Jun 5

1

Reposted by Quentin André

Stephanie M. Lee @stephaniemlee.bsky.social · May 29

Francesca Gino: "I am deeply sad and disappointed that the University has concluded as it has. But now that this process is over, I am free to show why its conclusion is so clearly wrong. I will continue the fight and do everything in my power to right this wrong." www.linkedin.com/posts/france...

From the start, I have said that I did not commit any academic misconduct… | Francesca Gino

From the start, I have said that I did not commit any academic misconduct and I want to reconfirm that statement. And from the very start, Harvard has blocked me from defending myself publicly. Firs...

www.linkedin.com

2 4 6

Quentin André @quentinandre.bsky.social · May 21

Two very skeptical birds on my hike today.

2

Reposted by Quentin André

Quentin André @quentinandre.bsky.social · May 15

Dear colleagues teaching methods classes,

Have you included any of my papers/blog posts in your syllabus? If so, I'd be super grateful if you could send me a short note, either here or via email.

Thank you!

1 2