Re: st: questions about using survey commands in Stata 8

[Steve Samuels <sjsamuels@gmail.com>]

I should have mentioned that subgroups within the age category of interest are also subpopulations. For these,  Stata 8's -svymean-, -svyratio-, -svyprop-, and -svytotal-  commands  probably had an over() option as do the equivalent commands today. But for estimation commands like -svyreg-, you will have to specify the full subgroup (your age group _and_  male, for example) in the subpop() option.

Steve



Have you  searched the -help- for Stata 8's survey commands or read the Survey Manual?
I don't have the manual any more, but I recall that the survey commands did have a subpop() option, for example:

****************************
svymean myvar, subpop(mypop) 
****************************

Steve
sjsamuels@gmail.com

On May 3, 2012, at 3:46 PM, Karen Rappaport wrote:

I am using Stata 8.2 to analyze a large public data base created from a survey. I would like to limit my analyses to subjects between the ages of 20 and 65. However, as I found on a website (http://www.ats.ucla.edu/stat/stata/faq/svy_stata_subpop.htm), I cannot simply delete all observations of subjects under 20 years of age and over 65 years of age. ("If the data set is subset, meaning that 
observations not to be included in the subpopulation are deleted from the data 
set, the standard errors of the estimates cannot be calculated correctly.") The information I can find indicates that I need to use Stata's subpopulation option(s).


However, the subpopulation options that are suggested in the link above can only be used in Stata 9 and 10. I do not know whether another subpopulation 
option exists for Stata 8 that I have not been able to find or if it 
simply does not exist.

Do I have no choice but to include all age groups in my analyses? How did 
researchers using Stata limit their analyses to subpopulations before 
Stata 9 was available?

I considered the possibility of creating dichotomous variables to separate the age groups I want to analyze from those I do not. However the "by" command does not work with survey data. (I want to analyze the subpopulation of people between the ages of 20 and 65 using other dichotomous variables, too.)

Please advise!

Thank you.
Karen

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/