1 paper
Andrew Gelman, Jessica Hullman, Lauren Kennedy
The average causal effect can often be best understood in the context of its variation. We demonstrate with two sets of four graphs, all of which represent the same average effect…