st: command cluster

["Michael Hunoldt" <michael.hunoldt@wiwi.uni-jena.de>]

Dear statlist users,

I am currently working on a study in which I apply the following regression command:

reg <dependent variable> objective_dummy, cluster(id)

My problem is that are multiple cases per id and that these different cases can vary regarding the value (1/0) of the dummy variable.

My question is: How does the Stata command „cluster“ handles cases in which the value of the dummy variable is not consistent? I would expect that State would calculate a mean value in these cases. Am I right?

I would appreciate any help.

Thanks

Michael