[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: RE: cluster option with fixed effects
In general, yes.
See the thread by started Jason Yackee:
and ends with Mark Schaffer's comments:
> -----Original Message-----
> From: firstname.lastname@example.org [mailto:owner-
> email@example.com] On Behalf Of Eliana Garces
> Sent: Thursday, November 23, 2006 10:28 AM
> To: firstname.lastname@example.org
> Subject: st: cluster option with fixed effects
> Dear all,
> I am running the following regression with fixed effects for countries and
> the cluster option.
> xi: reg lp totalq size i.subsidiary i.period au be de ge gr sw fi po uk
> hu sl lt lit lux ne no slo it if segment==0,cluster(country)
> Is it acceptable to use cluster for a variable already controlled for in
> the regression (in this case it is country)? I am getting extremely hight
> t-statistics (above 150 for some coefficients) that are much higher than
> the t-statistics of the regression without the cluster option.
> Thanks for the feedback.
> Eliana Garces
* For searches and help try: