Lol I had a BAD IDEA that i pitched to possible collaborators and they said "no professional upside to that idea Alex" and then i forgot to email you back with apologies and now im embarrassed in public.
Two half-baked thoughts (that might already be in the paper, sorry!!):
1) meta-analysis serves a "summary" function that might help us learn from the glut of papers
2) I think the "estimate-density" of papers will go up as competition increases (i.e. "metaketa" style work)
Congrats, what great news! Bravo!
Honestly, you should just change your handle to @andrew.37035397.phd !
are you advocating *silicon sampling*??? [ducks]
Friends, you've been submitting great proposals for this project and we're now coming up to capacity. We're closing proposals on April 15th; please send yours in!
Nice new entry in the persistence literature
🧵1/ Our first meta-science paper (with 350+ coauthors) is published today in Nature. It presents one of the largest-ever reproducibility projects in economics & political science.
Here’s what we found 👇
oh wow I need to read closer, 10 years ago!!
Re: your conclusion ("it depends on the research question") -- I still don't understand what RQ is answered by the estimator that controls for downstream things.
The fraction of the descriptive change in happiness not explained by the mediation model, supposing the mediation model is 100% correct?
From our FirstView article: What Happens When You Can’t Check the Box? Categorization Threat and Public Opinion among Middle Eastern and North African Americans by @asdurso.bsky.social. doi.org/10.1017/S000...
100% agree. be the change!
Godspeed! If it's of any use, political science wrestled with many of the objections that I imagine also come up with in sociology. www.dartstatement.org describes how our community handled transparency for different kinds of research. Not perfect, but these days, most poli sci papers have data!
Psychology has a whole cottage industry in which people come up with some construct that is essentially "attitudes/beliefs/expectations/feelings about X", and then the central claim is that this construct is a super important determinant of future X outcomes.>
I'm hiring a postdoc! @goetheuni.bsky.social
Focus: CSS, political behavior, political communication & transforming information environments.
📍 Frankfurt | ⏳ 3 years | 📅 Deadline: 14 April 2026
Full job ad here: www.uni-frankfurt.de/48794987/Zen... (search for “political behavior” to find it)
\{SOLID} latex joke Yamil!!
yes is awful dislike didn't want they did it to us
This project grew out of repeated failure to move party ID experimentally with "light" treatments (nulls that we document in this paper).
We learn from the hypotheticals project that HUGE treatments might change things (to the tune of about 5\% of partisans switching)
Scatterplot of switching
This scatterplot shows the correlation of pre-treatment party with hypothetical party ID; off diagonals show imagined "party switching". Not *zero* in the SQ condition, but substantially larger in the partisan hypothetical conditions.
SQ: For the next few years, the Russia–Ukraine war remains at a stalemate. Neither side advances far beyond where they are today. The fighting is constant but does not escalate. Pro-D Donald Trump wins the 2024 election and immediately withdraws all US support for Ukraine. Putin easily conquers Ukraine, slaughtering hundreds of thousands of people in the process. He then invades Poland, again killing many innocent people in his quest for domination. Pro-R Donald Trump wins the 2024 election and immediately negotiates a ceasefire between Russia and Ukraine. Russia withdraws from most of Ukraine, and hundreds of thousands of lives are spared because of Trump.
Here's an example of the status-quo, pro-D, and pro-R hypotheticals, after which people were asked what their 7 point party ID would be.
abstract: While attempts to change Americans’ partisanship via persuasive treatments largely fail, partisanship can and does change over time. In this paper, the authors first confirm, via survey and field experiments, that typical campaign messaging in the United States does not budge partisanship. The authors then present experiments in which participants encounter extraordinary hypothetical scenarios (e.g. one party causes economic collapse) before reporting what their partisanship would be under such circumstances. Twelve percent of partisans imagine switching parties in the pro-out-party hypothetical conditions, compared with 5% in the control hypotheticals in which the status quo persists, for a seven-percentage point (SE 1.5 points) difference. These hypothetical shifts are on par with the largest changes in American macropartisanship ever recorded. While the act of ruminating on hypothetical scenarios is not followed by changes in partisanship measured post-treatment, the evidence suggests that extraordinary world events may be able to shift partisan affiliation.
New paper with Don Green and @ethanvporter.bsky.social in the QJPS. After much deliberation, we went with a title that just states the result. 📝
journal: www.emerald.com/qjps/article...
I've been using the "three parameter selection model" developed by vivea and hedges as implemented here -- whats your take on that?
github.com/jepusto/meta...
also of interest @patrickpliu.bsky.social's thread on the working paper
Congratulations to @yamilrvelez.bsky.social, @patrickpliu.bsky.social, and @scottclifford.bsky.social !
we think attitudes are some function of beliefs; our exps routinely move beliefs but not (even correlated) attitudes. This team found a way to guess which beliefs matter more (and they do!)
P.S. Pre-post differences are *not* valid treatment effect estimates. Why? Here's a post by @statsepi.bsky.social: statsepi.substack.com/p/one-simple..., here's a post by me: www.the100.ci/2025/01/22/r... >
YES! that would be v. cool.
A clunky party of DD is the declare_model() section; if we could somehow do declare_model(daggity_spec) that would be amazing. it's just that all the details (the outcome spaces, strength of covariances) are hard to get in there simultaneously
Seems great!
!!!! will look into this, that sounds like it would solve many annoying things (like how it sometimes uses older code)
(should note that I checked the r2, it was wrong, I told the AI it was wrong, and it fixed it. So there were two rounds before I believed the sim.)
Screenshot of claude just writing a design no trouble
Writing simulations in DeclareDesign just went from "I should do that, but it's kind of a lot of work" to extremely easy