[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: F-statistic on ivreg2 vs ivreg with clustering
When I run the ivreg2 command with the "robust cluster()" options
(without the gmm option) I get a different Fstatistic than the one I
get with ivreg. Can anybody explain why?
In particular when I run
ivreg Dearn15 (rearn=age age2 eduyr eduyr2 sexo cavern) if filtro1==1,
I get an F( 1, 615) = 11.13
Prob > F = 0.0009
While if I do
ivreg2 Dearn15 (rearn=age age2 eduyr eduyr2 sexo cavern) if
filtro1==1, robust cluster(clust)
I get an F( 1, 614) = 1.60
Prob > F = 0.2068
I checked that this only occurs when the option "cluster" is included,
otherwise the output is the same under both commands.
This is not the result of gmm (because I didn't specify that option),
nor the result of any small sample correction.
If anybody with knowledge of ivreg2 can help me out why the output is
different under these commands I would greatly appreciate it.
* For searches and help try: