If you have a numeric relationship code which is small (e.g. one) for
the household head, you could:
. bys hhid (relationship): gen remit_hh=remit[1]
but you should answer Friedrich Huebler's question to your own
satisfaction before you proceed.
To do literally what you requested, you might
. gen notdad=relationship!=father
. bys hhid (notdad): gen remit_hh=remit[1]
assuming there is exactly one father per hhid.
There are equivalent methods using egen, e.g.
. generate fremits=remits if relationship==father
. egen remits_hh=max(fremits), by(hhid)
but these are less adaptable methods (IMHO)--
it's probably good to learn -bysort- sooner rather than later.
On 3/1/06, Friedrich Huebler <[email protected]> wrote:
> Marilyn,
>
> Out of curiosity, what kind of data has households in which everyone
> is either a father, mother, son, or daughter? What if there is no
> father or more than one father in a household? Wouldn't it be more
> precise if you had a variable that actually identifies the household
> head, which may be simply the person listed first in each household?
>
> Friedrich Huebler
*
* 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/