Cosal Inference

Inferring causation from observational data:

“The study is careful to account for all kinds of possible confounds, such as people’s characteristics and economic trends. The size of the effect is large on many dimensions: a 5 percent increase in a friend’s house value over the past two years corresponded to a person being 3 percent more likely to buy a house….”

Matt Jackson summarizing Social Networks and Housing Markets in The Human Network