Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Correctly identifying families in a household survey


From   "Deborah Garvey" <[email protected]>
To   <[email protected]>
Subject   Re: st: Correctly identifying families in a household survey
Date   Fri, 29 Jul 2005 14:59:05 -0700

Hi, list.

I've come up with a solution to the family identification problem that involves merging parent to child records, creating a new default family identifier based on the desired parent's value, and then assigning this value to the other parent via one of two merge commands.

I am happy to share my code with others.  Kindly email me privately.

Best,

Deb Garvey

>>> Deborah Garvey 07/29/05 10:06 AM >>>
Dear list members,

I have been struggling with a data manipulation problem and hope someone can offer a fresh perspective, as I'm getting nowhere.

I'm using individual-level data from the US 2000 Census, IPUMS version from the University of MN.  A household identifier (SERIAL) and a person locator within each household (PERNUM) together uniquely identify each person.  

Individuals related by blood or marriage should be in the same family unit (FAMUNIT), but sometimes they are not.
 
MOMLOC indicates the pernum of the person's mother; it is 0 if no mother is present in the household.  POPLOC is analogously defined.  These are the only linking variables between parents and children if FAMUNIT is incorrect.

Here's an example of the problem:  person 3 in the first household  should be in the same famunit as persons 2 and 4, since 3 is 4's mother and 2 is 4's father.  Similarly, person 8 in the second household is the father of person 9, but has a different famunit from his child and the child's mother (person 7).

        serial   pernum   famunit   momloc   poploc 
1.   2427039        1         1        0        0 
2.   2427039        2         2        0        0 
3.   2427039        3         3        0        0 
4.   2427039        4         2        3        2 
5.   2427039        5         4        0        0 

6.   3052039        1         1        0        0 
7.   3052039        2         1        0        0 
8.   3052039        3         1        2        1 
9.   3052039        4         1        2        1 
10. 3052039        5         1        2        1 
11. 3052039        6         1        0        0 
12. 3052039        7         2        0        0 
13. 3052039        8         3        0        0 
14. 3052039        9         2        7        8 

There are no family identifiers within households as long as FAMUNIT is incorrect.  How can I fix FAMUNIT so that it is always the same for related individuals?
 
I don't see a way to exploit positional relationships with explicit subscripts, since child(ren) can be listed at any position relative to a parent.   I can't use -bysort serial momloc- since there can be more than one 'mom' in a household.  I've tried merging the parents' FAMUNIT to the children, which was useful for determining the extent of the problem.  But, I can't see how to use the information to correct the parents' FAMUNIT values.

Any help is greatly appreciated.  I will keep plugging away..

Deb Garvey

******************************
Deborah Garvey, Ph.D.
Department of Economics
Kenna Hall
Santa Clara University
Santa Clara, CA  95053
408/554-5580
408/554-2331 (FAX)
[email protected]    
http://lsb.scu.edu/~dgarvey 
**********************************


This message scanned for viruses and SPAM by GWGuardian at SCU (MGW1)

*
*   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/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index