Unfortunately I couldn't make it to #SNE25 myself this year – but exciting to see our work up on stage from afar: Hilke Plassmann presenting our work in progress on how "The Skewed Distribution Of Likes Makes Social Media Users Less Happy" (with @robbrutledge.bsky.social & Klaus Wertenbroch).

Life satisfaction mostly declines with age. Previous findings (esp. the famous U-shaped age-SWB trajectory) were artifacts of misspecified models. doi.org/10.1093/esr/...

Made it across the pond :)

I am almost thirty, and my physical property amounts to about 3-4 suitcases. #academialife

This seems like a positive development and distinctly different from how Meta and TikTok have handled AI slop proliferation. It’s attempting to differentiate between spam content and AI generated creative content. techcrunch.com/2025/07/09/y...

Only a small % of people engage in toxic activity online, but they’re responsible for a disproportionate share of hostile or misleading content on nearly every platform Because super-users are so active, they dominate our collective impression of the internet www.theguardian.com/books/2025/j...

AFAIK there is no scientific evidence that supports a "neurological catastrophe of gigantic scale" related to social media. Giving platform to what seems entirely personal writing by a neurologist here is misleading. Can we tone down the rhetoric please? www.newyorker.com/magazine/201...

Last week I had the pleasure to present my research at the Marketing Department of Scuola Superiore Sant'Anna in Pisa on invitation of my friend and colleague Andrea Bazzani. Also really cool to see the Neuromarketing Lab that they are building there!!

"Bluesky is most definitely alive and kicking. CEO Jay Graber says the days of winner-take-all social networking are over." www.fastcompany.com/91342316/blu...

Mosseri (Instagram): "But you don't share personal moments in feed today the way you did five or 10 years ago. You share them in stories or in messages more so." - Seems about right. People now are more strategic and self-presentational about their timeline posts. www.lemonde.fr/en/pixels/ar...

We are pleased to announce the launch of the new Civility Analyzer module in @Communalytic.org. The module offers researchers an accessible, data-driven approach to studying both prosocial and harmful behaviours across various social media platforms. socialmedialab.ca/2025/06/17/c...

A recent Reuters survey finds that more people in the US get their news from social media than from any other source. I think that's mostly incidental exposure: You can hardly not read news on social media, whereas it’s an active choice to read the news in the newspaper.

In Dublin at @ucddublin.bsky.social for the 2025 #neuropsychoeconomics conference. Let me know if you are around and want to chat about #socialmedia x #decisionscience x #mentalwellbeing! www.neuropsychoeconomics.org/conference/

Political advertising on Facebook and Instagram before the 2020 US election had no effect. www.nber.org/papers/w33818

This summer, I will start as an Assistant Professor of Marketing at the #UniversityOfSouthCarolina Darla Moore School of Business. This is an incredible step forward for me, and it also marks the accomplishment of one of my life goals which was to become a professor b4 turning 30. Took a village :)

Is there starter kit / list for researchers in the #socialmedia space already? Would be nice to connect with people across #econ, #psych, #mkt and more...

Is there starter kit / list for researchers in the #socialmedia space already? Would be nice to connect with people across #econ, #psych, #mkt and more...

"How Germany’s AfD Harnessed Social Media. The far-right group has learned to communicate in a country where traditional media no longer shape public opinion." www.wsj.com/world/europe...

New paper on how a massive bot campaign is currently in action in Twitter/X to undermine the current german election... Spoiler : They better than the russian bots, and they're not afraid to spam 1000 tweets by day with IA help. PDF in english here (Canva link) : www.canva.com/design/DAGer...

We took X to court. And we won! ⚖️ The courts ruled that X must urgently grant us access to their publicly available data via their API. This access is critical for our analysis of online public discourse ahead of the upcoming German elections. Read: bit.ly/4gBWGp4 Here's what's happened 1/3 👇

🚨New WP🚨 Remember Musk+Zuck+Trump+Jordan etc crying fact-checker bias b/c Reps were flagged more than Dems? We analyzed Community Notes on Musk's X and guess what: posts flagged as "misleading" are 67% more likely to be written by Reps! The issue is Reps, not fact-checkers... osf.io/preprints/ps...

Here's a peek at what our @gruzd.ca & @philipmai.com have been working on this weekend: cn.deepfakestracker.org It's a new dashboard for tracking media on X that have been labelled as "misleading or manipulated" by Community Notes members. More info later this week.

PSA: Bluesky is a public social network, like a blog. Posts are visible to anyone online. Built on the AT Protocol, it supports PUBLIC, portable conversations by storing data in accessible repositories.

Quite the indictment of AI products: they are more appealing to people who don't understand how they work and see them as magical The paper concludes: "businesses may benefit from targeting those with lower AI literacy" and "maintaining an aura of magic around AI" ✨ tinyurl.com/ynefd8zk

[Breaking] Two Republican supporters of the TikTok ban in the Senate, Senator Cotton and Senator Ricketts, cast doubt on the notion that Trump will swiftly reinstate the app.

My first reaction to the #tiktokban.

Just a reminder for Germans abroad that you need to re-register to vote in the Federal Election, even if you didn't move houses since 2021 (yes, I'm a political scientist and had to google this). Here are the docs you need: www.bundeswahlleiterin.de/bundestagswa.... (1)

A sad tendency of many sci reform efforts is that instead of promoting transparency, certain research practices are just devalued entirely. Examples: 1. Exploratory research. Should be labelled as such but still valuable. 2. Correlative evidence. Does not establish causality but still valuable.

The "Free Our Feeds" campaign aims to protect social media from billionaire control by establishing a public-interest foundation. The initiative aims to ensure that the AT Protocol, the underlying technology of Bluesky, remains open and accessible to all.

I feel similar about social media btw Some of the posts on here (and elsewhere)... It's like road rage. Try reading them to yourself a week later, or better yet to a friend. In clearing out my Twitter account I saw some stuff from 'old me' that had these vibes, it's embarrassing!

Dear Tech Companies, No. We don't want "A.I." anything. We don't want A.I. social media "friends." We don't want A.I.-generated news reporters or anchors. We don't want A.I. to curate our music or make our movies. We REALLY don't want A.I. put in charge of our insurance claims or kids' education.

There's been a lot of discussion around Meta's decision to replace #factcheckers with #communitynotes in the US. Here's an attempted analysis on the consequences of this decision from a research angle. I am pushing back against some of the more pessimistic takes. 🧵

One of the reason I prefer Bluesky to X (as well as Threads/Facebook/LinkedIN) is simply because my newsfeed isn't controlled by an engagement-driven algorithm This reduces the incentives for gaming the algorithm with clickbait, conflict entreprenuers, etc. www.science.org/content/arti...

For folks using #Prolific, is there an option to report participants providing low-quality data? I prefer not to reject submissions. I wish there was a way to log low quality submissions and if enough people reported same participant they'd get booted from the platform

Meanwhile, Meta is preparing to launch bot profiles en masse on purpose. #platformdifferences www.ft.com/content/9118...

Textual features in the content that people post (e.g., anger-related words) help predict who is likely to share fake-news (versus, e.g., random or matched users). Matches my intuition, and other recent research showing the dark side of online outrage. journals.sagepub.com/doi/abs/10.1...

The word of the year is “brain rot” These days, it’s invoked by young people on social media to describe the “supposed deterioration of a person’s mental or intellectual state,” particularly stemming from overconsumption of trivial online content. www.nytimes.com/2024/12/01/a...

The limiting factor for MRI based prediction of behavioral traits may be neither the imaging nor the pipeline or learning algorith but rather the reliability of the target phenotypes Thought provoking work my Martin Gell et al: www.nature.com/articles/s41...

If you want to support our projects next year, please consider making a small donation: www.paypal.com/donate?campa... We are a publicly registered charity in Germany: www.educatgermany.com/impressum-pr...

A couple years ago, I founded a charity association: EDUCAT Germany e.V. Our goal is to support education projects in the Global South. 2024, we were able to co-fund the expansion of a community center in Buenos Aires, that will benefit 122 teenagers and children. www.linkedin.com/embed/feed/u...

Pleased to share the latest version of my paper with Arthur Spirling and @lexipalmer.bsky.social on replication using LMs We show: 1. current applications of LMs in political science research *don't* meet basic standards of reproducibility...

1) Random feed reduces toxic content 2) But also reduces platform use 3) And users more likely to share toxic content, mitigating the drop 4) And users with more interest seek it out/switch platforms Yeah, we might not be well equipped to deal with social media

In my experience, ChatGPT did a lot better generating code in JS and Python than in R. I guess this has sth to do with training data. Anyone aware of tools that do better at writing R?

"Politically neutral scientists are viewed as the most credible. [...] there is a monotonic penalty for scientists displaying political affiliations: the stronger their posts, the less credible their profile and research are perceived"

I'm compiling types of manipulations in psychology: 1) Vignette study 2) Autobiographical ("Remember a time you ...") 3) Confederate acts some way 4) Watch video inducing emotion 5) Behavioral instruction ("talk to a stranger") 6) Prime a construct 7) Framing outcomes To contribute, reply

🌍 Grüezi from Zürich! 🍫 🏔️ Today I had the pleasure to present my research at the brownbag seminar of the Technology Marketing Chair of ETH Zurich. Great to get feedback on our project and looking forward to continuing conversations. Thanks for the invitation!

Will be visiting ETH Zurich tomorrow – let me know if you are around and want to grab a coffee!! ☕

Data Viz Question: Say I have a hierarchical dataset with multiple observations (e.g. trials) from multiple individuals. I want to visualise the "average" *individual* distribution; i.e., how a typical distribution of all the observations by one individual looks like. What would you do here?

@sjdm-tweets.bsky.social in #NYC is a wrap! Four days packed with insights on everything #choice. Great to see many familiar faces, make new friends, and learn about a ton of cool research.