st: Error when estimating full model in -cmp-

[Jennifer M Keister <jkeister@ucsd.edu>]

Dear All,

I am having some difficulties running a multiequation model in -cmp-.

I have seven ordinal dependent variables (each with three possible values other than "missing") and six covariates for each (the last four equations have the exact same set of covariates, the first three have slightly modified versions of these covariates).

I have tried the following code:

cmp setup
cmp (DV1 = X1_1 X2_1  X3_1 X4_1 X5_1 X6_1) ///
  (DV2 = X1_2 X2_2  X3_2 X4_2 X5_2 X6_2) ///
  (DV3 = X1_3 X2_3  X3_3 X4_3 X5_3 X6_3) ///
  (DV4 = X1_4 X2_4  X3_4 X4_4 X5_4 X6_4) ///
  (DV5 = X1_4 X2_4  X3_4 X4_4 X5_4 X6_4) ///
  (DV6 = X1_4 X2_4  X3_4 X4_4 X5_4 X6_4) ///
  (DV7 = X1_4 X2_4  X3_4 X4_4 X5_4 X6_4), ind(5 5 5 5 5 5 5)

Stata runs the seven individual models without any problem.  However, when it moves on to fit the full model, I get the following:

Likelihoods for 1203 observations involve cumulative normal distributions above dimension 2.
Using ghk2() to simulate them. Settings:
    Sequence type = halton
    Number of draws per observation = 70
    Include antithetical draws = no
Each observation gets different draws, so changing the order of observations in the data set would change the results.

      cmp_lnL_censored():  3301  subscript invalid
               cmp_lnL():     -  function returned error
                 <istmt>:     -  function returned error
r(3301);

Many thanks in advance for any suggestions,

Jen Keister
Postdoctoral Research Fellow
Belfer Center for Science and International Affairs


*
*   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/