Dear colleague,
Arbitrarily, you can drop those one-member observations, or use dummy
to identify them. But this approach is not suggested.
In my view, it fits maximum-likelihood probit models with sample
selection. -heckprob- is a candidate for it. At the first state, you
esitimate why there is only one member in some observations. Then at
the first stage, you run the model to the observations with more than
1 member.
On 10/25/06, liu nizi <[email protected]> wrote:
Dear all
I want to build a logistic model to study the reason of some family
plunging in poverty.
I encounter some variables that represents characters (such as education
)of the second most contributing person to family income. The problem is
there are some family which have only one member can earn money. How can I
treat such variable?
Thanks
sarah
_________________________________________________________________
与联机的朋友进行交流，请使用 MSN Messenger: http://messenger.msn.com/cn
*
* 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/
--
Xiangping JIA
*
* 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/
