Re: st: using interactions with lags

[Scott Cunningham <scunning@gmail.com>]

On Sep 29, 2006, at 10:17 AM, Michael Hanson wrote:

> On Sep 29, 2006, at 9:44 AM, Scott Cunningham wrote:
>
> > Is it possible to use indicator variables interacted with lagged  
> > variables using within OLS?  For instance,
> >
> > xi:reg Y i.race*l(0/11)*X i.year, robust
> >
> > In other words, I want to interact the race dummy variable with 11  
> > lags of the X regressor.  I appear unable to do this using the  
> > above syntax.   SHould I, instead, move to generating the dummies  
> > manually and then interacting that with the lag?  In other words,  
> > is it the -i- operator that is causing the problem?
> 	One likely "problem" is with your notation:  L(0/11)*X is not a  
> valid expression for the lag operator.  Instead, I suspect you may  
> mean L(0/11).X, as in:
>
> xi:reg Y i.race*L(0/11).X i.year, robust
>
> See -help tsvarlist- for additional information.  I do not know for  
> sure if Stata will accept the combination of -xi- with the lag  
> operator, as I do not have a suitable dataset to test it.  Hope  
> this helps.
Dear Michael,

Yes - sorry about that.  It was a typo.  I did mean l(0/11).X, not l 
(0/11)*X.

I looked through the tsvarlist help, but unfortunately, there's no  
discussion about using the time series operators in combination with  
other operators, as far as I saw.


scott

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/