Ari Decter-Frain

@sean_reid @acerezo Awesome!

@ezagheni @MPIDRnews @d_alburez @monjalexander @Akbaritabar @ugobas I did this last year and can confirm it is a fantastic way for any demographer-in-training to spend a summer. Happy to chat with anyone interested in applying!

Call for the 2nd edition of the Summer Incubator @MPIDRnews open to PhD, master and advanced undergraduate students. Focus on forecasting and on crowd-sourced data. Mentors will include @d_alburez, @monjalexander, @Akbaritabar, @ugobas Fun and Funding will be available.

No

📢📢📢 Cornell Population Center is looking for another postdoc as part of our Rhodes Postdoc program! Offers lots of flexibility to focus on own work and collaborate with CPC affiliates. academicjobsonline.org/ajo/jobs/22698 Please share the word and apply! @PopulationCU @CornellBPP

@SebastianDeri @Google @GoogleWorkspace Congrats!!

I joined @Google as a Quantitative UX Researcher! I'll be working on understanding and improving (1) Search for Gmail, Drive, and Chat, (2) AI products for Google Meet, and (3) Google Forms. Grateful and excited to be working with the very talented team on @GoogleWorkspace :)

@george_berry This is fun! 8 > 5 > 6 > 1 > 4 > 7 > 3 > 2 > 9

@maria_antoniak I think this is my favourite advice post I’ve ever seen

The combination of attending NAACL, writing the acknowledgements section of my dissertation, and flying back to Ithaca for a final two weeks has me reflecting on a lot of things. I keep coming back to these "rules" and how helpful they were for me throughout my PhD.

Anyone making the intelligent design argument for the existence of God must explain why the Midwest is in the eastern half of the United States

@oscand_martinez @jccardenas1965 @marcfbellemare @nickchk @causalinf Might want to tag @grant_mcdermott ?

Hi #EconTwitter someone has any idea about ML for dummies related to environmental issues? Thanks a lot. @jccardenas1965 @marcfbellemare @nickchk @causalinf

Just met the big boss!

@george_berry @monjalexander has a paper with a stan model that combines observations from two different sources with different known variances. It's easy to tweak so that one or both variances are estimated. Repo is here: github.com/MJAlexander/fb…

@JaeJaeykim2 Coding Interviewer: And how could you make your solution more efficient? Me: add lazy_dt() at the beginning. Interviewer: you’re hired

@stevenstrogatz IRATE

Wordle players: What’s your favorite way to start? I often begin with STARE and COUCH. What two words do you like to open with?

@dmimno @_akpiper Can confirm

@_akpiper Yes, you definitely can

you really could spend a whole semester teaching topic modeling. understanding its behaviour across different document types, lengths, and model parameters. This would give you such a deep understanding of what "topics" capture.

@MarkDVerhagen @SociusJournal Very happy to see this out, Mark!

Although not the most joyous of times, I'm grateful to share that the first chapter of my DPhil has been published this week in @SociusJournal. Have you ever wondered why social scientists don't like to predict? So did I, and I find very little reason for this lacuna to persist.

Somebody give this guy his bag!

@lrberge should be rewarded, not punished for his efforts. And this package should really have the “Proudly supported by @rstudio” statement at the bottom of its docs by this point

This package is literally essential to my workflow. I think it is one of the best things R has going right now

@JaeJaeykim2 dtplyr to never change syntax though

@SebastianDeri @dggoldst This goes straight to the twitter HOF

@dggoldst The difference between the effect size reported in an original study v. the effect size in its replication

What's the prototypical strong, reliable effect in psychology?