Re: st: Right Censored ordered probit with Stata 11

[Austin Nichols <austinnichols@gmail.com>]

Urmi Bhattacharya <ub3@indiana.edu> :
Owen's program runs fine for me, and is far superior to -oprobit- in
this simple example:

drawnorm x1 x2 x3 e, n(10000) clear
g xb=(x1+x2+x3)/2
g y=(xb+e<-2)+2*inrange(xb+e,-2,0)+3*(xb+e>0)
g c=uniform()>.9
replace y=1 if c==1
oprobit y x1 x2 x3
coprbt y x1 x2 x3, censored(c)


On Fri, Sep 23, 2011 at 9:11 AM, Urmi Bhattacharya <ub3@indiana.edu> wrote:
> Hi Austin,
>
> Thanks for the putting the original code.
>
> This was the code I referred to ( though I should have put this up
> myself). It seems to me that the global macros like $ML_Y1, theta1,etc
> are not defined here which led me to think a part of the code is
> missing maybe. Or could be that I am going horribly wrong somewhere.
>
> Thanks
>
> Urmi
> On Fri, Sep 23, 2011 at 6:17 PM, Austin Nichols <austinnichols@gmail.com> wrote:
>> Nick--
>> Looks like the original code is here:
>> http://www.stata.com/statalist/archive/2003-07/msg00917.html
>> but I have never tried to estimate a censored ordered probit, and I
>> strongly suspect convergence will be difficult to be confident of.
>> It looks like Owen just adapted that original code into an ado for his
>> own use, and I don't see any further development,
>> but I am guessing Maarten Buis will have some input on this
>> problem--wasn't his dissertation research related to this problem?
>> (see what I did there?  we call that passing the buck in this country)
>>
>> On Fri, Sep 23, 2011 at 8:23 AM, Nick Cox <njcoxstata@gmail.com> wrote:
>>> I don't know what "the code provided by Edoardo" is.
>>>
>>> All I can see is the code in the Statalist posting your referred to in
>>> your first post. I see no obvious problems in the use of global macros
>>> there.
>>>
>>> Someone else may be able to help (much) more, but you'd increase your
>>> chance of better help by giving precise details of what you mean.
>>>
>>> Nick
>>>
>>> On Fri, Sep 23, 2011 at 11:03 AM, Urmi Bhattacharya <ub3@indiana.edu> wrote:
>>>> Thanks Nick.
>>>> I have been trying to use the code provided by Edoardo, but it seems
>>>> the global macros are undefined, suggesting a bit of the log file has
>>>> not been pasted. As for Owen Haaga's code, I can't figure out how to
>>>> frame the command after running the ado files defining the program.
>>>> I am hoping someone on the list has run  into a similar problem and
>>>> would be able to provide me with suggestions on how to work it.
>>>> Thanks
>>>>
>>>> Urmi
>>>> On Fri, Sep 23, 2011 at 12:47 PM, Nick Cox <njcoxstata@gmail.com> wrote:
>>>>>
>>>>> That posting included code, which you can copy. I have no idea how good it is.
>>>>>
>>>>> Owen Haaga hasn't been back to Statalist since, but the person with
>>>>> the same name who works at the Urban Institute may be the same person.
>>>>> Odds are that Austin Nichols will know.
>>>>>
>>>>> Nick
>>>>>
>>>>> On Fri, Sep 23, 2011 at 8:02 AM, Urmi Bhattacharya <ub3@indiana.edu> wrote:
>>>>>
>>>>> > I am interested in running  a right censored ordered probit for grade
>>>>> > attainment. I saw in the statalist archive that Owen Haaga had written
>>>>> > a generalized ado file for doing the same for version 8.
>>>>> > http://www.stata.com/statalist/archive/2003-08/msg00426.html
>>>>> >
>>>>> >
>>>>> > I was wondering how could I access the ado file  and run it ?
>>>>> >
>>>>> > Does someone know how if there is a different program for doing the same?
>>

*
*   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/