st: nlsur for different dependent variables in different subsample

[xjsunxm <sun.qqhr@yahoo.com.cn>]

I built a sample on network observations. According to our design, we need to
split the sample into two subsamples because we work on two types of
dependent variables. One subsample only contains observations of closing
structural holes with broker(1) and nonclosing(0), and the other only
involves observations of closing structural holes without broker (1) and
nonclosing (0). The two subsamples have the same observations on nonclosing
(0).

I try to run relogit model (more 0 than 1), using two equations (different
dependent variable and the same independent variables), one for each
'subsample'.i.e.,
      relogit y1 x1 x2 x3
      relogit y2 x1 x2 x3

Then I want to use nlsur to check the correlations among the errors, compare
the influence of Xi on two types of network change process (closing
structural holes with / without broker) and get more robust results.
However, so far as i know, the nlsur cannnot work on different subsamples.

A possible alternative solution is use 'suest'. i.e.,
      relogit y1 x1 x2 x3
      est Y1
      relogit y2 x1 x2 x3
      est Y2
      suest Y1 Y2
Again, the method suest does not work, and print out the error information
"suest requires that predict allow the score option
r(322);".


So, can anyone help me with the code I should be using for this? Any help
will be very appreciated!!!



--
View this message in context: http://statalist.1588530.n2.nabble.com/nlsur-for-different-dependent-variables-in-different-subsample-tp7565151.html
Sent from the Statalist mailing list archive at Nabble.com.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/