[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: RE: out-of-sample prediction of shares
I recently saw a talk by Phil Cross entitled "Partial identification of
conditional means and variances: using marginal distributions to sharpen
inference on mean regressions" which looks at this type of problem. His
contact details are available at http://www.georgetown.edu/faculty/pjc22/.
> -----Original Message-----
> From: Oleksandr Shepotylo [SMTP:firstname.lastname@example.org]
> Sent: Wednesday, August 27, 2003 4:39 AM
> To: email@example.com
> Subject: st: out-of-sample prediction of shares
> Hi statalist group,
> I would greatly appreciate any idea and/ or links to the journal articles
> (books) dealing with the (remotely) similar problem:
> I have two household surveys of the same population. Survey A has 5,000
> observations and detailed income data but
> is not regionally representative. Survey B has 50,000 households and is
> regionally representative
> but has no income data other than total income derived as
> income = expenditures + savings
> Data sets have a set of common variables X.
> My final goal is to predict income shares from skilled, unskilled labor
> capital in data set B
> based on estimation of income shares from data set A conditioned on X.
> So, my questions are:
> 1 Any suggestion on estimation procedure bearing in mind that dependent
> variables (shares of income
> from skilled labor unskilled labor and capital) are bounded between 0 and
> and add up to 1.
> 2 Any idea on better way to predict shares in the data set B based on
> estimations from A. (again the same limitations)
> * 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/
The information contained in this document is intended only for the
addressee and is not necessarily the views nor the official
communication of the Department of Labour. All final/official papers
which are sent from the Department will be sent by non-electronic
means, on appropriate letterhead, signed by authorised personnel.
* For searches and help try: