Brian Wansink's whole gig getting blown up by him admitting to p-hacking in a smug blog post is just an all time great self-own. Well worth reading about if you don't know the story. www.buzzfeednews.com/article/step... - ThreadSky

lastpositivist.bsky.social • 18 days ago

Brian Wansink's whole gig getting blown up by him admitting to p-hacking in a smug blog post is just an all time great self-own. Well worth reading about if you don't know the story.

https://www.buzzfeednews.com/article/stephaniemlee/brian-wansink-cornell-p-hacking

Comments

joesimons.bsky.social•17 days ago

Given it was titled "The grad student who never said no", I recall the general reaction at the time being "this is shocking - but not in the way you think!"

diacriticaz.bsky.social•18 days ago

This is hilarious but it's also a huge problem that scientists are incentivized to produce popular conclusions. We need more things like the Reproducibility Project, with bigger funding and most especially prestige. That is the most important part, you need to be able to make a name doing it.

ldnmike.bsky.social•17 days ago

We also need publishers to be interested in negative findings. There is a pressure to find positive outcomes to report, but the truth is most of the useful work in science is eliminating those things that don’t work.

gizmotronic.bsky.social•18 days ago

We also need dudes like Wansink to be shunned but …

diacriticaz.bsky.social•18 days ago

Yeah, but like this is kinda the problem. This is still a story about Wansink when it shouldn't be. I actually had to go look the whole thing up in Wikipedia to find the names Tim Van der Zee, Jordan Anaya, and Nicholas Brown, who should be the heroes here.

diacriticaz.bsky.social•18 days ago

* some of the heroes here, I'm only just finding out about all this and have to dig to find who did what.

diacriticaz.bsky.social•18 days ago

This does a better job but the whole thing is framed the wrong way. We should *completely erase* Wansink's name to the extent possible and we should be valorizing the people who caught him. https://www.thecut.com/2017/02/cornells-food-and-brand-lab-has-a-major-problem.html

diacriticaz.bsky.social•18 days ago

Notoriety is just another kind of fame, and fame is the real currency in which we pay academics.

jamesheathers.bsky.social•18 days ago

Hey, I'm so notorious you didn't even cite me.

(really, tho: investigations like this make other investigations easier... we were still really learning at the time, and some blossoms grew in the shit this case left behind https://sites.google.com/view/jamesheathers/home )

dcleclair.bsky.social•17 days ago

The way that pizza is cut creeps me out! What kind of monster does that?

jamesheathers.bsky.social•18 days ago

This picture of the pizza has pissed me off for YEARS. So rage-inducing.

The science bit, tho, you calm down after the first few years.

ldnmike.bsky.social•17 days ago

No shade but why is it always behavioural psychology? 👀

ghostingdani.bsky.social•17 days ago

I blame BF Skinner.

jsthrill.modphi.com•17 days ago

He couldn’t have known!

erikwestlund.bsky.social•17 days ago

I pronounce it "Phacking", no hyphen

Phacking and harking

aelkus.bsky.social•18 days ago

supervillain speeches are real

lastpositivist.bsky.social•18 days ago

oh hey this article's author is on BlueSky! I was inspired by her talk to post, @stephaniemlee.bsky.social

simoneskeen.xyz•18 days ago

This was a fun one because he was so obviously gunning for some fluffy pop-psych-airport-book-and-Oprah-guest trajectory before stepping on the rake. Not, e.g., derailing a generation of Alzheimer’s research.

jamesheathers.bsky.social•17 days ago

^ This is actually why I shifted my research focus. We could catch Wansinks all day. It wouldn't matter.

rich335.bsky.social•18 days ago

This will only get worse with AI because this is exactly how AI looks through datasets.

khakhalin.bsky.social•17 days ago

It's possible to correct for that though. Data exploration is a legit approach, especially to complex data. But there are ahem, checks and balances, that need to be invoked when you fine-grain like that.

cailinmeister.bsky.social•17 days ago

It’s a fab example for both pointing out that people often don’t realize they’re using bad methods, and that they propagate them via training @psmaldino.bsky.social

psmaldino.bsky.social•17 days ago

💯. The Wansink story is insane — the infinite soup rig! Also @stephaniemlee.bsky.social’s science reporting is 🔥

steamtraen.eu•17 days ago

Did you know that the bottomless soup bowl experiment was actually replicated? https://psycnet.apa.org/journals/xge/153/2/275/

(I was reviewer #1)

psmaldino.bsky.social•17 days ago

I did NOT know this. The first result about eating more soup makes total sense. The second result is sort of wild.

whaleactually.com•17 days ago

I miss the BuzzFeed news science desk so, so much

adw.bsky.social•18 days ago

I saw it happen in real time. It was just wild - “is this really happening?” vibes

lastpositivist.bsky.social•18 days ago

It's one of those things which if it happened in a film you'd think it was a bit much.

adw.bsky.social•18 days ago

The blog post was basically the villain explaining his plan to Bond while stroking a cat in a *parody* of Bond films 😂

huwgildon.bsky.social•18 days ago

The damning evidence is aways so much sillier in real life vs movies. An email that say, "Would you like to do crimes on these numbers, it will help?" "Yes I will do the maximum amount of crimes on these numbers I can do". It's not Ocean's 11 is it.

clarealice.bsky.social•17 days ago

Honestly I kind of believe him that he’d never heard of p-hacking or the replication crisis? It’s the only explanation that makes sense of how proud and open he was about it. Just totally incurious about how prominent people in his field are thinking about whether the results are true or not

archimandrite.bsky.social•17 days ago

Could be but then it points to something broken in the credentialing system for researchers.

gizmotronic.bsky.social•17 days ago

Is this clown still employed at Cornell? If so, what the hell are they thinking? If I had 15 papers retracted my university would probably disappear me.

lastpositivist.bsky.social•17 days ago

No he subsequently resigned

gizmotronic.bsky.social•17 days ago

I mean, I remember visiting in 2016 when his terrible approach was a known fact and indeed, I used his research as an example of basically unethical research conduct back then. And he was still there doing the occasional press release...

gizmotronic.bsky.social•17 days ago

It's coming back to me like a bad dream. The soup bowl thing was just made up. That's the thing, "oh I don't know about p-hacking." but also "oh, what? you can't make up data? what if its really good?"

votesforoysters.bsky.social•17 days ago

But still kind of baffling if he hadn’t. How did he manage that?

Comments

Posting Rules

Reply