in the case :
y
1
2
3
do you have two groups,
y group1 group2
1 1 1
2 1 2
3 2 2
but the groups not have the same number of components, in the survey case
the problem is similar. do you have two options
1. xtile y [fw = factor], nq(5)
or
2 do needs use the command expand, sort the variable, create an count, and
generate five groups.
Martin
> One way is to figure out the WEIGHTED cut points of the quintiles
> (1/5*sum of the weights) and create them yourself. Or one of the
> commands allow aweights, which will give you the correct point
> estimates, but not the correct standard errors.
>
> As always with percentile measures, be wary of ties, and consider if
> you will need to split weights to get exactly 20% in each group.
>
> Bryan Sayer
> Statistician, SSS Inc.
>
> -----Original Message-----
> From: [email protected]
> To: [email protected]
> Sent: 9/18/02 7:15 PM
> Subject: st: Surveys manipulation
>
> I am doing some research with a houshold survey, however, I'm facing
> the
> following problem:
> To generate consumption quintiles among the survey population first I
> sort the
> data according to consumption and then generate five groups. However,
> when
> weighting the data each group previously generated does not hold
> anymore the 20%
> of the population. Here comes my question, how can I generate quintiles
> in a
> survey taking into account sampling weights.
> Can anyone help me?
> Marcelo
>
>
> *
> * 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/
> *
> * 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/
*
* 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/
