st: Re: How to find out if there are different variables in differentwaves of survey data

["Michael Blasnik" <michael.blasnik@verizon.net>]

Sam Rawlings, Social Sciences and Law 03 wrote::
> I have data on the National Family Health Survey (NFHS) for India for two 
> waves of the survey - 92-93 and 98-99. The data I have is pooled for these 
> two waves and I have a dummy variable for each survey (named NFHS92 and 
> NFHS98) where the dummy var = 1 if the observation came from that wave and 
> = 0 if it did not (i.e. NFHS92 is the dummy for the 92 survey, and all obs 
> which are from 1992 have NFHS92=1).
>
> My question is this: I would like to know if there are some variables 
> which are present in one survey but not the other - and am wondering 
> whether there is a way this could be done in Stata.
>
> Currently I am just describing the data by the group NFHS92 -  this 
> results in a (very) lengthy list of vars and I am looking for cases where 
> N=0 as indication that that variable is not present in a particular wave 
> of the survey. Is there an easier, or perhaps more sophisticated way to do 
> this, with less output?
>
> Any help would be appreciated,
There are several ways you could accomplish this.  Here's one:

foreach var of varlist _all {
qui count if NFHS92==1 & !mi(`var')
local n92=r(N)
qui count if NFHS92==0 & !mi(`var')
local n98=r(N)
if `n92'>0 & `n98'==0 local only92 "`only92' `var'"
if `n92'==0 & `n98'>0 local only98 "`only98' `var'"
}
di "Variables only in 92 data: `only92' "
di "Variables only in 98 data: `only98' "

Michael Blasnik


*
*   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/