Ah yes, the good old "we tried chopping up the data many ways to find a specific three-way-interaction we had in mind". You love to see it in 2023.
@rubenarslan Ruben, you #statscop you! How dare you criticize your peers like this. Don't you realize that it is theoretically possible, with a non-zero probability, that trying out multiple ways of analyzing the data leads to the TRUTH? Surely you have to admit it is possible. So, from this it follows logically that this tweet makes you a horrible person for criticizing your peers. Do better.
@lakens @rubenarslan Yikes, where is this from?
@rubenarslan @lakens I see, oh well.
@rubenarslan
@lakens @shuhbillskee
Sorry, I'm not following. What am I missing?
The obvious "we wanted to find the IA, it wasn't there, so we tossed and turned the data again and again" aside, they do eventually conclude _no_ interaction and instead admit to the overall effect.
@GeorgKrammer @lakens @shuhbillskee if that's their modus operandi, how much do you trust the couple of other interactions they report in the same blog post?
@rubenarslan @lakens @shuhbillskee
Oh ok, just a #blog ...
#NoSource in original post :( only a pic of someone writing they tried finding something and eventually giving up. Could've even been purely #explorative
Who knows?
#Trust something on the #internet ? Oh come now... ofc DON'T! 😁
That's not new in 2023.
And always remember our good friend #xkcd
https://xkcd.com/386/
@GeorgKrammer @rubenarslan @lakens I can assure you these specific authors have a long history of not innocently exploring data in popular and academic writing. If you're curious about the specific I'd start with:
https://doi.org/10.1038/s41562-018-0506-1
Have a nice day. 🙂
I don't think there's nothing substantive to these trends, but, man, I don't trust that gang with their thirst for monocausal explanations to get to the bottom of this or anything else really.