Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: GMM bug when clustering on string var

Subject   st: GMM bug when clustering on string var
Date   Thu, 07 Jul 2011 16:20:23 -0500

Julian Reif <> pointed out that -gmm- with a string
cluster variable produces incorrect standard errors.

This problem will be fixed in a future update. For the time being, Julian
should use a numeric variable that encodes the clusters as he did in his

Here are some details drawn from his examples that illustrate the exact
problem and the work around.

Julian considered a regression of -y- on -x-. He defined -string_i- to be
the string variable with the cluster information, and -i- to be a numeric
variable with the same cluster information.  Julian then showed that

> gmm (y - {b1}*x1 - {b0}), instruments(x1) onestep vce(cluster i)

produced correct standard errors while

> gmm (y - {b1}*x1 - {b0}), instruments(x1) onestep vce(cluster string_i)

did not.

*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index