I have considered the maximization options. I do not wish to change the
iterate() or ltolerance() in particular (in fact, when I increased the
ltolerance to 1e-1 nothing changed (still flat region error) despite the
iterated log likelihoods being < 0.1 apart). Reducing the number of
iterations does result in the model completing with an output, although many
estimates and 95% CIs are missing for variables. I am not sure how changes
to the other options would help ³push² the model to give me estimates.
-Andy
*
* 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/
