5 ms·
> "Remarkably, the friends became ill about two weeks before the random undergraduates, probably because they were, on average, better connected. With the world
by rdlecler1 10y ago
> "Remarkably, the friends became ill about two weeks before the random undergraduates, probably because they were, on average, better connected. With the world only imperfectly prepared for a pandemic, being able to spot trends in this way could be useful."
Isn't this just because you have a larger sample (N people * Y friends) than if you are looking at just N people? In which case it's more likely that one's friends will get sick before the randomly selected person. If so what's suprising about that?
Does anyone have a link to the original paper?
- pizza 10y agoI think it's saying that peoples' friends are better predictors of illness than just people.
- rdlecler1 10y agoBut isn't that just a function of the larger distribution that gives you more resolution on the tail? I assume I'm missing something here because that seems obvious.
- hkhall 10y agohttp://journals.plos.org/plosone/article?id=10.1371/journal.pone.0012948 http://journals.plos.org/plosone/article?id=10.1371/journal....