The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Weighting at regional level (Weighting )
Weighting at regional level [message #26413] Fri, 17 March 2023 11:42 Go to next message
seinkim is currently offline  seinkim
Messages: 3
Registered: October 2022
Dear DHS team,

I'm using calendar data of Zambia 2018/2019 and do analysis on contraceptive use. I want to aggregate contraceptive use per month and region, and do the event study. The outcome variable is rate of condom use/pill use/injectables/etc per month at particular region. What I did to produce these outcome variables at that particular month was to aggregate numbers of contraceptive users divided by total women surveyed at that month. In this case, how should I use weighting? Could you give me some advice?

Thank you!
Re: Weighting at regional level [message #26431 is a reply to message #26413] Mon, 20 March 2023 11:28 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2742
Registered: February 2013
Senior Member
Following is a response from Senior DHS staff member, Tom Pullum:

Are you aggregating with the "collapse" command in Stata? Sometimes I find it convenient to use "collapse" to calculate a weighted numerator and a weighted denominator , bypassing a weight option, and then doing a division in a collapsed file. The following lines suggest a framework:

gen unwtdn=1
gen wtdn=v005/1000000

collapse (sum) *wtdn, by()

I don't know whether this will apply to your situation. Let us know if you need more detail.

Previous Topic: All-women factor in trend analysis
Next Topic: Weighting variable Madagascar
Goto Forum:

Current Time: Mon Mar 27 18:41:15 Coordinated Universal Time 2023