julia kamin
@juliakamin
connecting research to practice @ prosocialdesign.org. supporting social cohesion @ civichealthproject.org. dog mama in NYC.
Mind blown. Just like us humans, bacteria have an innate sense of fairness and punish group members who don't play by the rules. (According to Damasio in The Strange Order of Things.)
What's better than a review article? A review of review articles! This one is on interventions for cyberbullying and other online harms. journals.sagepub.com/doi/pdf/10.1...
Am late in finding this 2025 study showing that prompting Instagram users to share in others' positive emotions while scrolling had a significant impact on their wellbeing (d=0.25). As much as watching a 10 min video about dam-building beavers. :-) static1.squarespace.com/static/550b0...
Why are bridging conversations effective at reducing toxic polarization? One theory is that they break down our stereotypes by defying our expectations of what a typical counter-partisan believes. A new preprint - cleverly using chatbots - lends evidence that is the case. arxiv.org/pdf/2606.15901
Someone studied the landscape of Bridge-building initiatives in the U.S., including @listenfirstproject.bsky.social's coalition of 400+ organizations, and created a map.
Prediction markets - Kalshi and Polymarket - now have their money (just slightly) on the Dems taking the Senate. That was a far-fetched dream six months ago.
Love a review article. This one is of studies from 2020-2025 looking at AI's ability to detect cyber-bullying. link.springer.com/article/10.1...
A positive sign from new study: In survey experiments, participants were less likely to engage with content when they knew it was AI generated. link.springer.com/article/10.1...
We've reached the point when more than twice as many Americans Strongly Disapprove of Trump as Strongly Approve.
Had a similar thought "who believes a word he says?" But we have to give the markets some credit; I'm guessing the signal they are picking up is *not* that a ceasefire in the works, but that Trump is desperate for one. That small signal seems to be reflected in the relatively small adjustment:
There are more global users of Whatsapp than Facebook, but fewer than 5% of studies on digital media and politics looked at Whatsapp. A new paper makes the case for why that needs to change. www.ojcmt.net/article/soci...
New study & promising approach to *keeping* college students open-minded. www.tandfonline.com/doi/pdf/10.1...
Someone did a mapping of the US Bridgebuilding field. Cool. @listenfirstproject.bsky.social @civichealthproject.bsky.social onlinelibrary.wiley.com/doi/pdf/10.1...
He's been trending in that direction for a bit (I keep looking for when the number of strong disapprovers is double that of strong approvers), but I hope the Yougov isn't an outlier - as the Silver Bulletin composite suggests it could be:
New study: Researchers in Germany tested a version of "correcting misperceptions" to see if it could mitigate hostile attitudes between the far left and far right. I could. onlinelibrary.wiley.com/doi/pdf/10.1...
New study: AI chatbots can be effective at reducing polarizing attitudes - especially when directed to exhibit receptiveness and active listening. www.researchgate.net/profile/Mark...
Great review from @emilykubin.bsky.social and @vaparker.bsky.social on the challenges of researching what effectively reduces polarization - including knowing if and when reducing polarization should be something we want to do. www.researchgate.net/profile/Emil...
An RCT across 120 villages in Bangladesh suggests that film can be a powerful way to increase inter-group trust - especially if viewed by folks who are socially centrally connected. academic.oup.com/qje/advance-...
A reminder that Covid is not the only thing we should be testing - and staying home - for. 3,500 New Yorkers were hospitalized with the flu last week they counted - compared to 450 with Covid. And it doesn't look like we've hit the peak.
For any who think about both online child safety and polarization in the U.S. This report from More in Common shows strong cross-partisan support for digital protections - and suggests that it's an issue that help bridge divides. moreincommonus.com/publication/...
Any chances Cuomo had of winning today have got to be shot by the ballot design. Only determined Cuomo voters are going to put the work in to find his name in a sea of independents and may be forgiven for getting confused when they they think he's running on the Protect Animals ticket w/ Silwa.
13% of Americans now think polarization is our most important problem, while 15% of Americans think our most important problem is either Republicans or Democrats. We're divided about why we're divided.
Years ago, when I lead a talk series in NYC had tons of conversations with folks with extremely strong political views. I remember thinking how - regardless of whether they were from the left or right - they all seemed eerily similar. Maybe this explains in part why. www.apa.org/pubs/journal...
An example of where GenAI gets it wrong. There's been so much written about the use of Ranked Choice Voting in the NYC *primaries*, GenAI's modeling imagines the general election uses RCV too. It does not.
3% of Americans are satisfied with the release of info on the Epstein files. I don't think I've seen 3% on any poll. Ever. You can usually count on inattention or error to get you to at least 7% (which is, btw, how many Republicans are satisfied). Astounding. www.cnn.com/2025/07/15/p...
Of the many telling things from Mamdani's win yesterday is how much the Bronx and Brooklyn have diverged politically, with Brooklyn Dems (predictably) backing Mamdani over Cuomo 49% to 32%, but the Bronx swinging in the other direction, 34% to 53%. www.nytimes.com/interactive/...