Re: st: Could someone help me with my nl evaluation program? I update my program.

[张志强 <zhiqiangzhangnk@gmail.com>]

Thank you Nick, I will take into account your suggestions about
spatial dependency. May be the spatial econometric approach would be a
good choice for these kinds of regression.The endogenous variables
problems would also render the   consistent and robust of  related
parameters.
the NL2SLS methods would be a good choice.
My project has a long way to go!

Thank you very much!


2014-02-02 Nick Cox <njcoxstata@gmail.com>:
> I've separated out some questions here.
>
> Nick
> njcoxstata@gmail.com
>
> On 1 February 2014 14:54,  Zhiqiang Zhang wrote:
>
>> Thanks for your advice, I learn  a lot about the nl programming in stata.
>> I final got my nl estimation result for my naive wage equation. The
>> problem is because of my data, I found there are some errors in my
>> dataset. As for my program,  does it can be used  for much more
>> complicate summation terms?
>
> Difficult to comment, but as already pointed out in this thread, more
> complicated models tend to be difficult to fit.
>
>> DO I need to use the bootstrap to get much
>> more accurate estimation results for complicate project? such as the
>> distance term is a 2700*2700 matrix?
>
> Bootstrap at most gives more honest standard errors. If your problem
> is intrinsically spatial, you may have problems with dependence
> structure that render the bootstrap dubious any way.
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/