[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: ivreg2 and panel group heteroscedasticity
The latest version of ivreg2 supports panel data by requiring the data
to be tsset. But how does it interpret the _iis_ (individual identifier)
variable? Suppose there is no autocorrelation but group-wise
heteroscedasticity. Does the _robust_ option give the right standard
errors or they only correct for arbitrary heteroscedasticity? Will
cluster option with _bw(1)_ do the trick? Moreover, to use _cluster_ , I
have to mean difference my data ( to remove individual unobservables).
As noted in Wooldridge (2002), Chapter 10, do I correct for the degrees
of freedom in the standard errors?
Any response will be much appreciated.
* For searches and help try: