Our findings reveal that news media companies have an incentive to produce news headlines containing out-group derogation to maximize engagement, which may, in turn, heighten animosity between groups.

This was led by @jesperrasmussen.bsky.social @steverathje.bsky.social @crobertson500.bsky.social

6/6: Comments more than welcome! osf.io/h3cn5_v1

5/6 This research builds on previous research which has focused on the effect of identity language (onlinelibrary.wiley.com/doi/full/10....) and negativity (www.nature.com/articles/s41...) in isolation. Our research suggests that combining these two are even more powerful!

4/6 We also provide a causal test of the “out-group animosity” hypothesis: Our estimates suggest that negativity about out-groups (e.g., political opponents) boosts engagement by 14.1%, while negativity about in-groups has a much smaller effect (1.6%).

3/6 Headlines that include both negativity & identity language increase clicks by ~9.9% — significantly more than effect of negativity (4.3%) or identity language alone (5.9%).

2/6 We examined thousands of headline experiments from the Upworthy Research Archive (upworthy.natematias.com). This unique dataset allows us to test the causal effects of negativity + identity language.. Paper by @natematias et al. www.nature.com/articles/s41...

2/6 We examined thousands of headline experiments from the Upworthy Research Archive (upworthy.natematias.com). This unique dataset allows us to test the causal effects of negativity + identity language.. Paper by @natematias et al. www.nature.com/articles/s41...

Great discussion and sorry for jumping in. In my reading these examples (by @brendannyhan.bsky.social @aecoppock.bsky.social & Andy Guess) illustrate @hugoreasoning.bsky.social point. I think I've also seen similar evidence by @ethanvporter.bsky.social @thomasjwood.bsky.social (maybe I am mistaken)