The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Tanzania » Testing for the pool-ability of DHS data
Testing for the pool-ability of DHS data [message #5601] Mon, 15 June 2015 07:18 Go to next message
Robert is currently offline  Robert
Messages: 19
Registered: April 2015
Location: Dar Es Salaam, Tanzania
Member


Hi All

I am trying to pool DHS data sets for the three round up for Tanzania (DHS 1999, 2004 and 2009)
Could anyone with the idea on how to do the pooling? How do I make the strata survey round specific? Any one with a do file (or commands)
Regards
Robert


R.Lihawa
Re: Testing for the pool-ability of DHS data [message #5627 is a reply to message #5601] Wed, 17 June 2015 17:28 Go to previous message
Reduced-For(u)m
Messages: 292
Registered: March 2013
Senior Member

Robert,

As for "how to pool" - use the "append" command. It is really easy:

use DataSet1.dta
append using DataSet2.dta
etc....


As for creating new variables, there are lots of ways. One way would be to add a "round" ID when appending, and then use a simple, if ad hoc, method like this:

use DataSet1.dta
gen round =1
append using DataSet2.dta
replace round=2 if round==.
etc....

Then you can do something like this:

gen strata = round*1000 + strata

You can do something very similar for cluster, you just need to multiply the "round" variable by a number with as many 0's as there are digits in the strata/cluster/whatever variable you want to make unique.

If you have multiple countries, you can do something similar by generating a country ID number (in the appending process) and then multiplying that by a number with an extra 0 or two (so that your final ID would be "country number" "round number" "strata number" concatenated together.

That make sense?
Previous Topic: Use of Pooled Data set for PHD
Next Topic: Computing the U5 mortality, infant and neonatal mortality
Goto Forum:
  


Current Time: Thu Mar 28 04:40:54 Coordinated Universal Time 2024