LLMs have been proposed for annotation tasks. But, LLMs are biased and make errors. Can we draw * valid * conclusions from LLM annotations? arxiv.org/abs/2408.15204

Data science for economists: version control, web scraping, spatial analysis, workflow, project management and other needs. github.com/uo-ec607/lectu…

*Very* excited for this paper-led by amazing Cameron Martel (on the job market!)-to be out. He validates a scale for trust in fact-checkers & presents experiments w 14k subjects showing that fact-checker warnings reduce misinfo belief + sharing even among those low in trust!

My paper is out in PNASNews! I replicate a paper on the impact of COVID vaccine mandates on vaccine uptake. Removing a single bad control variable sign-flips several of the paper’s headline results. The reply’s findings are also not robust. 1/x pnas.org/doi/10.1073/pn…

🚨Out in Nature!🚨 Many (eg Trump JimJordan Elon Musk) have accused social media of anti-conservative bias Is this accurate? We test empirically - and it's more complicated than you might think: conservatives ARE suspended more, but also share more misinfo nature.com/articles/s4158…

The field should reward effort to understand boundary conditions - not treating boundaries as enemies. A good micro-founded model helps characterize when a theory works, and more importantly, why it might fail.

Great article from Kai Kupferschmidt about the five biggest challenges facing misinformation researchers highlighting some of our recent work! Agree with pretty much all these. science.org/content/articl…

.Gordon Pennycook et al. found that while psychological inoculation helps detecting misinformation at scale globally, it may not be effective unless it is combined with accuracy reminders. @STWorg David G. Rand @dgrand.bsky.social Adam Berinsky Hause Lin nature.com/articles/s4156…

In light of the past week, this recent paper by Adam Przeworski merits a re-read. The argument that “democracy’s defenders must offer a forward-looking perspective that seeks to improve representative institutions” is particularly relevant. Free download: muse.jhu.edu/pub/1/article/…

New WP! The illusory truth effect (repetition -> belief) is core to psych of beliefs, & thought to be a deep bias impacting misinfo, persuasion & advertising Why would cognition include such a flaw? We argue it is a rational adaptation to high-quality info environments 🧵1/

Economist Joshua Gans uses o1-pro to generate a (minor, fun) paper in an hour based on an idea of his, and it gets published in an appropriate peer reviewed journal, with adequate disclosure. He ends with the same sentiment I am increasingly seeing from fellow academics: what now?

Can emotion regulation interventions spread from treated to non-treated people in a group? We examine this question in a paper published in Nature Communications, led by Michael Pinus & Yajun Cao. nature.com/articles/s4146…

An experiment on RUMORS in NETWORKS. We find even for messages with no emotional or political content (Heads or Tail), people like to: 1. Overshare 2. Care about who they are sharing with. People are good at using the network as filter. papers.ssrn.com/sol3/papers.cf… #econtwitter

Remember this study about how LLM generated research ideas were rated to be more novel than expert-written ones? We find a large fraction of such LLM generated proposals (≥ 24%) to be skillfully plagiarized, bypassing inbuilt plagiarism checks and unsuspecting experts. A 🧵

Experimental validation, which links economic preferences to choices in incentivized elicitations, may fail to produce reliable new measures, from Jonathan Chapman, Pietro Ortoleva, snowberg, Leeat Yariv, and Colin Camerer nber.org/papers/w33520

Fascinating look at the political implications of AI and partisan differences by Beatrice Magistro Sophie Borwein R. Michael Alvarez Bart Bonikowski Peter Loewen in AJPS doi.org/10.1111/ajps.1…

I'm extremely proud of this new paper, out Association for Psychological Science, and extremely fortunate to have worked on it with the inimitable Amanda Geiser and Deborah Small. We find that when comparing moral wrongs, people are (much) more willing to "scale up" than "scale down" condemnation...

Back to X just to post new PNAS paper for Elon Musk: we find Community Notes flags 2.3x more Republicans than Democrats for misleading posts! The issue is Reps sharing misinformation, not fact-checker bias... pnas.org/doi/full/10.10…