That elderly man has a cousin fourth removed who once worked for the trash department of the Gaza municipality, which was ran by Hamas, and that's why our soldiers raped him to death in Sde Teiman

the "i can excuse x" meme really shining nowadays

A few months ago I read somewhere that he wants to naturalise 1948 Palestinians; haven't heard about it elsewhere or since, tho.

yeah germanness sometimes clashes with self-respect

not yet german enough to use a store bought sauce tho so will have to make my own which is very annoying

is hollandaise THE sauce for asparagus?? i thought i needed to boost my germanness yesterday so i bought a pack

syrian social media doing its thing

💯. I have been here for almost 10 years and I'm no longer able to deal with the dissonance of enjoying my peaceful life here when the same government (and society) making this life possible is wilfully and unrelentingly involved in the mass murder of thousands of people. Fucking sickening.

Would you say that weighting (by the joint distribution of the variables involved in the confounded interaction) is a viable solution in this case to recover population ATE?

fucking called it www.reuters.com/world/middle...

I hate that I have to think about this scenario, but say they agree to a ceasefire and release the hostage(s). What's stopping Israel and US going "psych, you have to do this other stuff" and continuing the genocide a few weeks later like last time?

I just offered the same idea to a friend when thinking about how Jolani can entice Trump: Trump Tower in old Damascus.

hmm i'm not sure what the implication of this is. one would fit a different model depending on whether an interaction is assumed right? or is the position of people who'd say that you should fit an interaction anyway?

one representing an assumption that there is an interaction, and another that doesn't, right?

i have a vague idea what it means but don't immediately make the connection to interactions. i also understand why DAGs don't NEED to represent interactions but imo if we want a reasonably precise representation of a hypothesis then it's probably good to differentiate between two DAGs, -

because most people don't know what "parametric" means???

i was gonna ask whether it wouldn't be too sacrilegious to instead represent the interaction effect as its own node and it turns out this has already been suggested (obv) pmc.ncbi.nlm.nih.gov/articles/PMC...