I can't advise on conditional logit versus alternatives.
The data management side of this question is to exclude
singletons. For that, calculate the frequencies and
then insist that frequencies must exceed 1.
bysort parent : gen freq = _N
... if freq > 1
Nick
[email protected]
somsupa Nopprach
> Thank you Prof. Maarten Buis very much for your
> suggestion.
> I just realized that some of my lines were be
> cut.Sorry for my mistake.
>
> For example, I have data like this.
> Parent firms Capital Affiliate Country Year
> AAA 233 aa Thailand 1997
> AAA 100 ab Thailand 1996
> AAA 200 ac Malaysia 2000
> BBB 50 ba Philippines 1995
> CCC 80 ca Indonesia 1996
> and so on.
>
>
> I would like to analyze the location choice
> determinants of parent firms' repeated investment
> decision.
>
> I thought that I should run conditional logit
> and use option-group(Parent firms. By the way,I do
> not sure that this method is correct or not.
>
> In addtion, if I would like to not include data that
> does not have repeated investment like Parent firm
> "BBB". How I can limit sample range in conditional
> logit model?
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
