Huge thank you to the co-author team: @stepalminteri.bsky.social, @livia-tomova.bsky.social, Johannes Eichsteadt, Aadesh Salecha, Shashanka Subrahmanya and, most importantly, the incredible supervisor trio of @orbenamy.bsky.social, @georgiaturner.bsky.social @lukasgunschera.com

Read the preprint below:
osf.io/preprints/ps...

This work brings together two lines of research: (1) evidence that social needs may be homeostatic, similar to hunger or thirst, and (2) findings that social media posting can be modelled using a reinforcement learning framework: an approach that can be used to answer key research questions.

In a preregistered study of 535 US Twitter users, we found that people learn more from social rewards during periods of high isolation. That is, when isolated, they weigh 'Likes' more heavily in deciding how often to post.

This is incredible, huge congratulations - so well deserved. 🎉🎉