Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: really need help please - fractional probit unbalanced panel-Wooldridge/ Wagner


From   Ken Ning <kenscholar2008@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: really need help please - fractional probit unbalanced panel-Wooldridge/ Wagner
Date   Wed, 12 Jun 2013 09:00:31 +0100

Hi Kelvin,

Yes my response variable fractional.

On page 24 of Wooldridge's slides, he said "xtreg allows unbalanced
panels and properly computes standard
errors and test statistics."

he also used coding such as
xtreg  math4  lavgrexp  tobs3_lavgrexp  tobs4_lavgrexp  lunch  lenrol
y95  y96 y97  y98,  fe  cluster(distid)

time dummies can only be included as time averages as the model has a
fixed effect.

after reading his slides, i am slight confused as how to use xtreg to
do the test.

thanks,
Ken






On 12 June 2013 07:58, Kelvin Mulungu <kelvin.statalist@gmail.com> wrote:
> Ken,
> Am not sure plain xtreg will handle your response variable which I
> have assumed is a fraction or proportion.
> Make it clear for me; Is your response variable fractional?
>
> If yes, and unbalanced data. Your considerations remain between glm
> and the user written Stata programme by Woodridge (frac_het) (as in
> the earlier link). But in this case, you are assuming unobserved
> effects. Time dummies can be included.
>
> Am not necessarily an expert here, but its something I have been
> learning equally.
>
> On Tue, Jun 11, 2013 at 7:03 PM, Ken Ning <kenscholar2008@gmail.com> wrote:
>> Hi Kelvin,
>> thanks- it seems Wooldridge suggests to use xtreg for unbalanced
>> panel- am I right?
>> how did you add fixed effect to the model? simply add time average variables?
>> it would be great if you could provide some examples
>> thanks
>> Ken
>>
>> On 11 June 2013 16:22, Kelvin Mulungu <kelvin.statalist@gmail.com> wrote:
>>> Am also interested in your first part.
>>>
>>> Let me try to refer you to a document that may help to aswer your
>>> second question about unbalanced data. Am assuming your response
>>> variable is a fraction, hence the choice for the fractional category
>>> of models.
>>>
>>> In this paper
>>> http://stata.com/meeting/chicago11/
>>> Woodridge explains how to handle an unbalanced panel with fractional
>>> probit and provides a user written Stata programme to run the same
>>> model.
>>>
>>> Kindly refer to the link for more details. I have personally used it
>>> and it worked, but I have not been able to find the interpretation.
>>>
>>> Thanks.
>>>
>>>
>>> On Tue, Jun 11, 2013 at 2:23 PM, Ken Ning <kenscholar2008@gmail.com> wrote:
>>>> anyone knows about this?
>>>>
>>>> On 10 June 2013 23:26, ken first <kenscholar2008@gmail.com> wrote:
>>>>> Dear Stata user,
>>>>>
>>>>>
>>>>> I really need your help with these two questions.
>>>>>
>>>>>
>>>>> Question 1. determining fixed or random effects in a fractional probit
>>>>> panel model
>>>>>
>>>>> Papke Wooldridge 2008 and Wagner (2008)suggests if the panel is
>>>>> balanced, one can use xtgree  or glm .
>>>>>
>>>>> for example,  xtgee dependent_variable independent_variables, fe
>>>>> family(binomial) link(logit) vce(robust)
>>>>>
>>>>> is this correct?
>>>>>
>>>>>
>>>>> I dont know how one determines if one should use a fixed or random
>>>>> effect in this case. Can one still use hausman tests to decide?
>>>>>
>>>>>
>>>>>
>>>>> Question 2. unbalanced panel
>>>>>
>>>>>  Wooldridge later on said in this post,
>>>>> http://www.stata.com/statalist/archive/2012-05/msg00585.html
>>>>>
>>>>> he has worked on some solution to unbalanced panel. Does anyone know
>>>>> what I should use if I have unbalanced panel data?
>>>>>
>>>>>
>>>>> Many thanks
>>>>> Ken
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> reference
>>>>>
>>>>> Leslie E. Papke, Jeffrey M. Wooldridge 2008 Panel data methods for
>>>>> fractional response variables with an application to test
>>>>> pass rates
>>>>>
>>>>>
>>>>> Wagner, Joachim (2008) : Exports and firm characteristics: first evidence
>>>>> from fractional probit panel estimates, University of Lüneburg Working
>>>>> Paper Series
>>>>>
>>>>> *
>>>>> *   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/
>>>
>>>
>>>
>>> --
>>> Kelvin
>>> kelvin.statalist@gmaill.com
>>>
>>> *
>>> *   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/
>
>
>
> --
> Kelvin
> kelvin.statalist@gmaill.com
>
> *
> *   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index