Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: t statistics instead of z statistics in a large sample ?.

From   Dahlia El-Hawary <>
Subject   st: t statistics instead of z statistics in a large sample ?.
Date   Sat, 16 Apr 2005 18:35:22 -0400

Could someone please tell me why do I get the t-statistics reported 
instead of the z- statistics in a large sample of over 50,000 
observations (6812 firms, 37 countries, 7 industries and t=4,...16) 
when I estimate a dynamic panel data model using GMM difference or GMM 
System estimators using xtabond2 command ?.

I do not specify small option in the estimation equation. 

I am suspecting that this (reported t- statistics) has to do with the 
WARNING message that I get when using the GMM difference or GMM System 
estimators in xtabond2:

 "Two-Step estimated covariance matrix of moment conditions is 
singular. Number of instruments may be large relative to number of 
groups. Using a generalized inverse to calculate robust weighting 
matrix for Hansen test". 

The reason why I am linking the warning message to the reported t-
statistics is that the largest number of instruments I ever had is 306 
instruments while the number of groups(firms) is 6,812. So it cannot be 
that the number of instruments is larger than the number of groups ?!. 

So, may be there is something wrong with STATA identifying the 
dimensions of my data set ?. I dont know ?. When I tsset my data, the 
panel id and the time variable are correctly identified.  

I would very much appreciate your help.
Thanks in advance,

Dahlia El-Hawary

PhD Candidate
Economics Department
George Washington University

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index