I want to understand when and why multi-way clustering is necessary.
Could anyone give a good example that provides intuition of when and why to use multi-way clustering that also illustrates why clustering at a group level is not sufficient?
$\begingroup$
$\endgroup$
2
-
$\begingroup$ Paper: Robust Inference With Multi-Way Clustering Leaving this for anyone else who stumbles across this page looking for information on clustering. I will also provide an outline of the paper once I have a bit of time this week. $\endgroup$– 123Commented May 2, 2016 at 12:33
-
1$\begingroup$ Also, you might want to have a look at the empirical limitations of clustering: economics.mit.edu/files/11355, in other words: when is clustering a bad idea. $\endgroup$– dv_bnCommented May 2, 2016 at 14:09
Add a comment
|