
So exciting to _finally_ get to talk about the US 2020 FB and IG Election Project! A few notes from my perspective as a methodological collaborator kind of on the border between the academic and the Meta teams. 🧵

Prof. of Data Science for the Common Good @Hertie_DSLab // @nyuniversity Politics PhD // former @Meta Core Data Science, @ds_vienna // he/him

So exciting to _finally_ get to talk about the US 2020 FB and IG Election Project! A few notes from my perspective as a methodological collaborator kind of on the border between the academic and the Meta teams. 🧵

Some professional news: after nearly ten years at The Washington Post, I'm super excited to have finally started something of my own: Broome Street Labs.

AI evaluation is all the rage these days. I'm excited to be teaching a class on it this Fall; I wrote up how I'm thinking about it: ddimmery.com/posts/ai-eval-c…

The Hertie School Data Science Lab is hiring a postdoc. This is an appealing position with fairly low teaching/admin responsibilities and a lot of freedom to develop your research agenda. Feel free to reach out if you want to know more! hertie-school.dvinci-easy.co…

Next a foray into refreshing and updating the workings of my website and a reflection on why this is Good: ddimmery.com/posts/website-r…

I'm back to blogging on my own pixels. I should remember to actually share posts on the microblogs (I believe this is the main thing they're good for). First, a post laying out why I won't be blogging on Substack: ddimmery.com/posts/back-to-b…

RCTs are good, but not (just) because they make the statistics easy. open.substack.com/pub/drewdi…

The US2020 Project was not independent (and that's good) open.substack.com/pub/drewdi…

How The Washington Post’s forecast will track 2024 election results Like in previous cycles we'll be running our live election model to add context to results that are coming in. But we have a few updates this year! washingtonpost.com/politics/…

Welche Koalitionen könnten nach den Wahlen in Sachsen und Thüringen anstehen? Zusammengefasst: in Sachsen könnte es sehr schwer werden, eine Regierung ohne BSW zu bilden - und in Thüringen beinahe unmöglich. Auf @zeitonline mit simulierten Wahlausgängen: zeit.de/politik/deutschland/…

Welcome #ICML2024 to Vienna! If you work on causality/CSS and want to chat, hit me up. I'm also happy to share recommendations based on a few years living here.

Our polling average topline takeaway: Trump is leading in 5 of the 7 battleground states that are most likely to determine the outcome of the election, but PA, WI and MI are within a typically-sized polling error. w/ @dianesaysword, @kt_prry & Luis Melgar washingtonpost.com/elections…

To assert democratic control over large online platforms we need rigorous impact assessments like the US2020 project. This is the only way we can make sure we respond to the real societal harms they may cause.
Do Facebook, Instagram and co. lead to polarisation? @DrewDim debunks this myth: deactivating accounts before the 2020 US election had little effect on political polarisation. People who logged off were even less likely to believe misinformation. 🔗 bit.ly/3z8Pj8K

New article from the U.S. 2020 Facebook & Instagram Election Study in @PNASNews (academic team led by @HuntAllcott & Gentzkow): Deactivation of FB/IG for 6 weeks before the 2020 election reduced online participation but otherwise had mostly null effects pnas.org/doi/10.1073/pnas.23…

***Academic Job*** I'm hiring a post-doc (starting this fall) as part of building out a lab model re: conflict and development at Emory. Open broadly but Asia is bulk of agenda right now. ABDs & recent grads please submit apps by June 15th apply.interfolio.com/145540

Tired: constant marginal effects Wired: maximum marginal effects at p=0.5 Inspired: add a parameter for where marginal effects are maximized Are they not teaching the kids about scobit these days? Embarrassing