
Nathan Hoffmann
@nathanihoff
PhD candidate in sociology @UCLA and incoming AP @emory_sociology researching immigration, education, sexuality, and quantitative methods. He/him š³ļøāš
ID: 1849566715
https://nathanhoffmann.com/ 09-09-2013 21:53:26
2,2K Tweet
1,1K Followers
1,1K Following


I've been extensively using Vincent Arel-Bundock {modelsummary} R package for both teaching and research and it is a delight. But I'm here to recommend this presentation of {marginaleffects}, a beautiful approach to interpreting model estimates: youtube.com/watch?v=ANDC_kā¦



Our Inequality Data Science Research Lab summer party, which fell on the hottest day of the year! Ian Lundberg Nanum Jeon Nathan Hoffmann Kristin Liao Shiva Rouhani



Happening today: Iāll be presenting new work with Nathan Hoffmann on geographic distribution of immigrants in same-sex couples. OPR posts records on their website afterward.


I am thrilled to share that I will be joining Emory Sociology as an Assistant Professor this fall! Excited for a new chapter in Atlanta š„š


Very pleased to have Nathan Hoffmann joining us in the fall! #soctwitter #sociology #emory #emorysoc


My latest article with Nathan Hoffmann forthcoming at Public Opinion Quarterly | How Sexuality Affects Evaluations of Immigrant Deservingness and Cultural Similarity: A Conjoint Survey Experiment osf.io/preprints/soca⦠via Center for Open Science

Come see me present new work with Kristopher Velasco from our project on queer migrants at #PAA2025!



When I first tried GPT3 I remember thinking writing as a skill was basically over. Now I think the opposite is true. LLMs still canāt write, everybody elseās abilities are atrophying, and being able to clearly express yourself in prose seems like it will only get more valuable

Cool paper from MIT Media Lab now showing clear proof of cognitive debt when using AI to write essays. Prior work from Vishakh Padmakumar , Max Kreminski had shown how AI assisted writing leads to homogenization. Over reliance on AI for cognitive tasks clearly has its share of harms.



wacky that āai writing detection is impossibleā has seeped into the popular consciousness all while theres strong proof that Pangram Labs can detect it i have no financial affiliation w them, they are just 1 of 2 subs that i am actively grateful for (other one being midjourney)
