Suppose the sources I'm learning from start using echo chambers of their own, making the high-quality sources more accurate and the low-quality sources less accurate (as above). Then I'm more inclined to want to filter out the low-quality sources by using an echo chamber myself. (4/n)

So echo chambers can bring those close to the truth closer, but also prevent the misinformed from learning much. We show that this can make echo chambers contagious. (3/n)

Echo chambers can increase the accuracy of beliefs on average, since people are more willing to heed information they trust. However, those who start with really inaccurate beliefs may be worse off using echo chambers, since the sources they trust are also misinformed. (2/n)

Some airlines may have bereavement fares for family, but either way condolences

The rational agent paradigm is often maligned as unrealistic, with experimental evidence of seeming irrationality cited as support. But this new work argues that people just can’t absorb or remember all the information that’s thrown at them, and have to make choices about what to focus on.

I’m guessing the increased engagement for low quality sources is not just people quoting them to explain their shortcomings?

It’s a lot of work to get to the bottom of why you disagree with someone—hard to think it’s worth it, when they’re probably just misinformed, or arguing in bad faith

So for example, maybe you don’t want to criticize your party’s candidate in front of undecided voters. But among friends, you’re willing to admit you have doubts about their ability to win, or don’t love all their policies.

But there are other good ones, too. Echo chambers can provide people with similar beliefs or preferences a place to credibly share information they might otherwise have misrepresented in attempt to persuade others.
See doi.org/10.1016/j.ge... and olejann.net/wp-content/u...

That’s the tradeoff we study in this paper: arxiv.org/abs/2010.01249

Sure there is. Conditioning your participation on the policies of the candidate you prefer gives them an incentive (your vote) to pursue policies you like.

Sure anytime

(proof: if the widest parts are at different heights, then you save space by alternating. And if not, you don't use any more space.)

For the US, FRED is a good place to start for macro data, including survey data on consumer expectations e.g. fred.stlouisfed.org/series/MICH

The existence of free will may have moral implications (is it right to lock somebody up for a crime if they couldn’t have chosen not to commit it?). But free will or not, people do tend to respond to incentives.