Florian Coste-Chareyre

patrick star from spongebob squarepants is holding a notepad and writing on it ALT: patrick star from spongebob squarepants is holding a notepad and writing on it

Nice! :)

Oops! Well, good luck ;)

Ty for the reference! :)
Its hard to keep track of all the modeling options with non-linear SEM, let alone understand how they work « under the hood »…

(Bollen & Pearl, 2013)

From Pearl’s SCM/SEM framework, a lack of path is clearly a stronger assumption than a present path. In linear models, a lack of path implies cov(x,y)=0, while a present path implies cov(x,y)=\0

But for my current issue, multi-level data with lots of non-linear effects, Bayesian estimation was the best option, at least I believe

Oh u are actually working on SAM! In fact I read your paper on non-linear capabilities of SAM and was considering using it in future studies!
Pretty neat if SE are available now 😎

(I meant to say =3)
I suppose now there is the newly developped SAM approach that is better at handling lots of non-linear terms, irc.

What are you referencing as not true ? I suppose my statement is conditionnal on the size of the (latent) control variables set. At >=3, if you incluse all 2-way interactions and the one 3way interaction, you are already at the maximum LMS can handle properly according to Muthén B, irc.

a close up of a dog 's face with its eyes closed and looking at the camera . ALT: a close up of a dog 's face with its eyes closed and looking at the camera .

Accurate

Very low. If you go SEM: u’ll need enormous sample size and to go bayesian estimation since ML won’t handle it

Well, in principle, yes, non-linear controls should be included. But if your controls are measured with error (as it can often be the case in psychology)… Good fucking luck 😂😂
If u stay in regression setting: it won’t adjust for anything because the reliability of 2nd and 3rd order terms will be >

Error term: « I am a joke to you ? »

This is Ollie. He found a perfectly placed carrot snack on his walk. Simply can't believe his luck. 13/10 best day ever (TT: maxime_aries)

This is Tuffy. He just woke up from surgery. Has no idea where he is or what's happening, but he's thrilled to be here. 12/10 (TT: teddy_and_tuffy)

On the good book, of course

🎂 Happy birthday to the European flag! 🇪🇺

🗓️ In 1985, it became the EU’s emblem, but did you know it was first adopted 70 years ago today by the Council of Europe?

Learn more about the EU flag facts and figures ➡️ link.europa.eu/cxYjT4

This! Admitting budget/time contraints + sensitivity power analysis would not only be more honest but also more usefull for the reader than this « apriori » power analysis nonsense

I agree - bayesian estimation is often a pragmatic choice when MLE simply doesn’t converge in complex model situations. I wonder how many users of bayesian stats are truely « philosophical » users !

Although, tbf, for the total predictive validity (and not only the incremential one) of that manifest to manifest best case scenario, one would have to add .57 * .41 * .67 * .62 = 0.10. So .16, which still is pretty bad!

Yeah, this looks pretty bad…

Poutine : "Je veux très littéralement détruire l'Europe".

(Une partie de) la gauche : TOUT CELA N'EST QU'UNE MANOEUVRE POLITIQUE DE MACRON.

Les gens qui nient la menace russe, je les vois comme ceux qui niaient le danger du COVID en pleine pandémie.

Combine*, rather

U can use both with the model-implied instrumental variable (MIIV) framework 😎
scholar.google.com/scholar?hl=f...

It is surprising I agree. Also I wonder if some samples (e.g. n=140) can actually handle (well) 2nd order factor models

There still seems to be a lot of confusion about significance testing in psych. No, p-values *don’t* become useless at large N. This flawed point also used to be framed as "too much power". But power isn't the problem – it's 1) unbalanced error rates and 2) the (lack of a) SESOI. 1/ >

The cherry in top is the fact that the relashionship between 2 of the variables are likely bidirectionnal

a man in a suit and tie is sitting with his eyes closed . ALT: a man in a suit and tie is sitting with his eyes closed .

The more I learn about causal inference, the more I wonder if choosing a thesis’ subject involving a complex moderated mediation model and psychological variables that can only be nudged through soft intervention was the right call… 🫣

Why are guinea-pigs the colours they are? Sewall's got a DAG for that.

We couldn't not rate this classic early DAG by our beloved collective granddagy, Sewall Wright.

14/10. Ten for the DAG, plus one for each cute guinea-pig node.

From Wright (1920) "The Relative Importance of Heredity and Environment in Determining the Piebald Pattern of Guinea-Pigs"