With a new semester comes a new ECR Forum program!🎉 PhDs and Postdocs at CIS are invited to attend. Buket Demirci (@ETH_en) and I look forward to seeing many of you there and send a big thank you to our speakers !

Check out our program below🗓️

How does misinformation affect the spread of disease? According to our model misinformation could have led to 47 million additional COVID-19 infections in the U.S.! Check out our paper, just out on the arXiv Matthew DeVerna (@mdeverna on bsky) Francesco Pierri Observatory on Social Media yy IU Luddy
arxiv.org/abs/2402.11351

journals.sagepub.com/doi/full/10.11…

Great paper from Magda Breyer @mabreyer.bsky.social: '[...] The results show that a stagnating representation trajectory mobilizes women to vote for a progressive party [...].

In contrast, [...] men do not lash back against women’s increasing representation [...]'

If you wonder what tokenization means about misinformation on this (or any other social media) platform, check out this study by Meysam Alizadeh, Emma Hoes, and Fabrizio Gilardi
nature.com/articles/s4159…

Very happy to share this new pre-print!

Christian Pipal and I explore how politicians employ TikTok to communicate with society:
doi.org/10.31219/osf.i…

We have an open position for a phd candidate / doctoral researcher at the Computational Social Science Department at GESIS in Cologne, Germany: hidden-professionals.de/HPv3.Jobs/gesi…

PhD Scholarships to work with me (on AI & human behavior) or many other scientists in the Max Planck Society school for cognition Max Planck School of Cognition

cognition.maxplanckschools.org/en/application

🚨Job alert!🚨

TWO Assistant Professors in Public Policy.

Join our wonderful department SGIA Durham and research group Centre for Institutions and Political Behaviour.

Link: jobs.ac.uk/job/DBS916/ass…

Happy to answer questions over email!

New pre-print on LLMs for text classification, co-authored with Youngjin (YJ) Chae

We systematically evaluate different approaches to text classification using LLMs, from zero-shot learning to models fine-tuned on large corpora.

osf.io/preprints/soca…

1/5

🚨 Just out in @nature Scientific Reports🚨

What is the type of content people would share on social media if they were rewarded for it with 💰💰💰?

In this paper, Meysam Alizadeh Fabrizio Gilardi and I show that even a simple nudge (not yet actual money)... [1/X]

NEW PREPRINT: Our hands-on guide for web browsing data! Led by Bernhard Clemm w/Sebastian Stier Ana S. Cardenal Ericka Menchen-Trevino Magdalena Wojcieszak

Covers analytical decisions re:
- Filtering/cleaning
- Measurement
- Content classification
- Experiments
- Modeling

LINK: osf.io/krhmq

⚡️Job alert ⚡️

2-year postdoc position, part of my PRODIGI
project IPZ - Political Science UZH

> Digital tech & politics
> Causal inference or computational social science
> PhD completed or close to completion
> Starting Jan 2024 (or later)

Happy to answer any Qs!

fabriziogilardi.org/resources/jobs…

If you find that study interesting, also check out our companion paper 'Open-Source Large Language Models Outperform Crowd Workers and Approach ChatGPT in Text-Annotation Tasks'

with Meysam Alizadeh Maël Kubli et al.

arxiv.org/abs/2307.02179

#ChatGPT #LLMs #OpenSource

Very happy that our paper 'ChatGPT outperforms crowd workers for text-annotation tasks' is now out at PNASNews 😀

with Meysam Alizadeh & Maël Kubli

pnas.org/doi/10.1073/pn…

#ChatGPT #LLMs

We're pleased to announce the call for papers for the 2023 ESOC Annual Meeting. Submit by June 15th to be considered: esoc.princeton.edu/events/2023/ca…

🚨 New Pre-print! 🚨 psyarxiv.com/zmpdu

Do interventions against #misinformation help, or do they come with undesired effects by increasing skepticism in otherwise trusted pieces of information?

Both! Our findings suggest current strategies require substantive changes 🤯

Interested in coalition governance and lawmaking? 🤝📚 David Schmuck, U. Sieberer, and I explore when and why coalition governments mandate inter-ministerial collaboration on legislative projects. Read the #OpenAccess article in Governance Journal: doi.org/10.1111/gove.1…

I was coming home from my Easter break last night and decided to play around with the ChatGPT API to pass the time 🤓. While I had used the text-generator before, I was baffled by what I found when I asked the AI to generate images... 🧵

Can ChatGPT be leveraged to assist fact-checking efforts? Yes! We find that ChatGPT accurately categorizes 12.000 fact-checked statements in 72% of cases. It is better at identifying true claims (80%) than false claims (67%).

Sacha Altay 🏝️ Fabrizio Gilardi AI, Media & Democracy | @[email protected]