As any good psychologist, I barely ever read anything written by people outside of my field and have no idea who most of them are. But the first time I read something by Manski I immediately knew that he must be an absolute legend, and he hasn't disappointed so far. - ThreadSky

dingdingpeng.the100.ci • 22 days ago

As any good psychologist, I barely ever read anything written by people outside of my field and have no idea who most of them are. But the first time I read something by Manski I immediately knew that he must be an absolute legend, and he hasn't disappointed so far.

Comments

tkaiser.science•22 days ago

I saw Pearl referencing him for his bounds on causal effects. That's a concept largely unknown to psychology, too, AFAIK.

dingdingpeng.the100.ci•22 days ago

Now that I did bounds for age/period/cohort effects, I want to put bounds on everything

dingdingpeng.the100.ci•22 days ago

I hope they give him a Nobel Prize at some point just so that I can react with "Oh, Charles Manski? They should have given it to him ages ago, he's great!"

jeremylabrecque.bsky.social•22 days ago

Yeah but then we can’t make the joke every year that the bounds on him winning are 0-1.

dingdingpeng.the100.ci•21 days ago

You were making that joke? Without me???? 🥺

jeremylabrecque.bsky.social•21 days ago

I feel like I see it every year around Econ Nobel time and it never gets old.

wildestreasure.bsky.social•22 days ago

Isn't it only logical and ethical for assignment of responsibility to be contingent on or established after valid proof and execution of correct modelling? The fact that it has to be spelt out so explicitly worries me.

timpmorris.bsky.social•22 days ago

Can’t claim to have read Manski in depth, but have skimmed his stuff attacking MI and not that impressed. The attacks on MI apply to most other methods of analysis with incomplete data, and the actual alternative to using these methods is to give up and go home.

eckles.bsky.social•22 days ago

Well he often argues for and develops methods for using weaker assumptions, and yes that often only results in set identification

timpmorris.bsky.social•22 days ago

Perhaps I’m being unfair and should read in more depth?

dingdingpeng.the100.ci•22 days ago

No I think you're right that it applies to other methods as well! And the thing that I do know him for is partial identification, so you move away from point estimates to bounds. I like that approach, although I can see why it's not more popular.

solomonkurz.bsky.social•22 days ago

Hmm. If the main point is "Hey guys, did you know you could cut yourselves with a kitchen knife?" I'm not interested.

dingdingpeng.the100.ci•22 days ago

I don’t think it’s as simple as that!

zetaof1.bsky.social•22 days ago

Well, honestly, sometimes every body needs a good reminder of what tools can do and what not. Scientific insight has been re-discoveres all throughout history.
For example, the argument that made Edmund Gettier famous was already made by Russell half a century before yet nobody cared

timpmorris.bsky.social•22 days ago

Ah ok. I should read more before saying more, but my knee-jerk is to be a bit skeptical. Anyway, I won’t say more now and will try to read more…

dingdingpeng.the100.ci•22 days ago

If your knee-jerk reaction is to be a bit skeptical, I feel like you should actually like Manski 😂

Comments

Posting Rules

Reply