[Limdep Nlogit List] skip
William Greene
wgreene at stern.nyu.edu
Tue Aug 11 02:48:49 EST 2009
Phil. The RP estimator is skipping observations with missing values
as it goes, but it is not changing the global observation count. You
can produce the same result, with the appropriate report of the number
of observations by using REJECT before fitting the model. You will have
to recompute the group size variable in that case, however.
/Bill
----- Original Message -----
From: "Philip A. Viton" <viton.1 at osu.edu>
To: limdep at limdep.itls.usyd.edu.au
Sent: Friday, August 7, 2009 1:17:18 PM GMT -05:00 US/Canada Eastern
Subject: [Limdep Nlogit List] skip
For panel data models for binary choice we're advised not to try and
manage missing data. I'd always taken that to mean, not to turn SKIP
on. But there's a problem: at least for the rpm models, you don't get
the banner announcing how many observations have been skipped, and
the "Number of Observations" line in the results is wrong (it reports
the current sample size). Turning Skip on seems to solve the problem,
but given the advice, I think it would be nice to be reassured.
So, question: assuming that -999 indicates missing data, are there
any circumstances in which turning SKIP on will lead to incorrect results?
------------------------
Philip A. Viton
City Planning, Ohio State University
275 West Woodruff Avenue, Columbus OH 43210
viton.1 at osu.edu
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au
More information about the Limdep
mailing list