Alex Coppock

Since it’s the #equinox, time to re-post the best #dataviz I’ve ever done: 365 pictures of the Earth from space, taken at exactly 6:00 a.m. Details on the NASA Earth Observatory: earthobservatory.nasa.gov/images/52248...

Congratulations! Would love a 🧵 on this one!

Guided by our values and theory of policy design, Searchlight’s approach to policy will be defined by bold, persistent experimentation.

this part seems cool

I wish someone would rewrite my crappy extreme value bounds / trimming bounds package for attrition to have variance estimates and tidy / glance methods

love this #notamap as a way of understanding the scale of land use. we're so used to seeing state borders and we have a rough sense of their relative sizes -- this plot makes great use of that

Ooh I especially dislike when an author preregisters a bad analysis (like something that conditions on post-treatment variables) then says in response to reviewers that they can't remove it because PAP.

glad to know about the rigorous treatment of this "commitment bias" behavior!

love this -- so very useful.

btw you *can* write your own handler as you did here (which is v. flexible and extensible) but you can also do many post-estimation things with the .summary argument if you like!

The great Vincent Arel-Bundock shows how to use marginaleffects post-estimation summaries with DeclareDesign simulations!

Lol hello elevator friends!

VAB has it right on both counts.

1. I ALMOST always describe the model in terms of POs, then calculate the estimand from them -- but you can also do a "structural model"

2. DD is happy with any DGP functions that make a data.frame, so if they make missing data...

Grateful to Matt Baum, Jamie Druckman, @jonmladd.bsky.social @brendannyhan.bsky.social @dannagal.bsky.social for joining the Author Meets Critics session on The American Mirage tomorrow at 12:00. If you spot me at the conference, ask for my favorite book merch: a custom magnetic bottle opener!

yep, there's a bit of popping up and down...

I'm once again a session chair for APSA and will be imposing the Nyhan / Samii rules. Encourage your chairs to do the same.

I don't see any good arguments for the usual

all papers
all discussant comments
all questions

format

wouldn't that be great?

methods classes are usually all about "tools" to answer research questions and only rarely mediations on the unanswerability of many research questions

whaaaat!? that's so cool!

literally just:

weird_trick = D * X

iv_robust(weird_trick ~ D | Z)

to get the mean of X among compliers!

All statistical models are wrong, but some are stupid

Suppose there are only two publication tracks for an experiment:

1) Pre-registered experiment:
PAP -> Realization -> Submit -> RR -> Publication

2) Registered report:
PAP -> Submit -> RR -> Realization -> RR2 -> Publication

when should we want 2 not 1?

to protect against "null risk"?

Interesting -- was the claim that ppl could sniff out a deepfake but now they can't?

I'm souring on registered reports lately... but I love testing pipelines on simulated data!

Trouble is my simulated data is usually nice and clean... and it's the asymmetric *cleaning* that often flips a sign!

"no hope of making scientific progress" is classically Mungerian, but I don't think he's right. If the value of some parameter varies over time, then we make scientific progress if we measure it repeatedly. He says academia is "too slow" for this; I think we just don't like to repeat experiments.

It is *so hard* to fight this kind of asymmetry. Any suggestions or workflows you do to counteract?

Notre Dame’s Department of Political Science is hiring *multiple* senior folks (assoc/full prof) working on democracy broadly (institutions, political behavior, CP, AP).

I’m not on the committee (hiding on research leave), but I'm happy to chat / answer questions!

👉 apply.interfolio.com/172986

This post seriously conflates "polling" with "survey experimentation"

Also any paragraph that starts this way is a big 🚩

LESS looks like A LOT of fun -- now you've got one more reason to be happy you live in London or one more reason to be jealous of those who do!

Submitted!

My sister is taking her post-tenure (WOO!!!) sabbatical trip to Nepal and will be blogging it!

I

Thank you Fabian! this is the first time i'm seeing it used by others, makes my day!! (plots look great!)

Really excited about this project. Also, shoutout to @aecoppock.bsky.social for the vayr package. Many folks really impressed by the elegance of his sunflower adjustment for over-plotting points.

alexandercoppock.com/vayr/

Thank you Zander! Let's meet up IRL!

Thank you to the Yale staff for all their support, especially Pam Greene for organizing (among other things) two book conferences for me!

Thank you to the students in my research design, experiments, and dataviz courses, without whom I could not have learned the material.

Come visit Evanston!