The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Pooled country analysis weighting
Pooled country analysis weighting [message #22425] Thu, 11 March 2021 04:55 Go to next message
shafaq.mahmood93 is currently offline  shafaq.mahmood93
Messages: 4
Registered: March 2021
Member
Hi,

I am trying to do a pooled data analysis for 6 south-Asian countries (Pakistan, Afghanistan, Bangladesh, India, Maldives and Nepal). I am trying to look at the trends of contraceptive usage and fertility preference of south-Asian Couples in the last 30 years.

Can you please guide me how to account for weights for each survey. Do I need to redefine the weight variable? Secondly, can you also guide me how to deal with the strata variable (v022).
Re: Pooled country analysis weighting [message #22432 is a reply to message #22425] Thu, 11 March 2021 15:46 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3017
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

There have been many postings on using "egen group" to distinguish the clusters and strata in the different surveys and construct svyset. I won't repeat here.

I recommend that you look at trends in each of these countries separately, and then compare them. You CAN pool all the surveys from all these countries but there will be lots of heterogeneity. If you look at each country separately, you can stay with the original weights. If you combine countries you will have to alter the weights, or else the large surveys and countries (especially India) will completely dominate the small countries (especially Maldives).
Re: Pooled country analysis weighting [message #22441 is a reply to message #22432] Fri, 12 March 2021 03:47 Go to previous messageGo to next message
shafaq.mahmood93 is currently offline  shafaq.mahmood93
Messages: 4
Registered: March 2021
Member
Thankyou for your guidance.

As you have suggested that if we combine all the countries then we will have to alter the weights. Can you please share any guidelines on how to alter the weights? We have a senior statistician in our team who has a sound understanding of all these things. But we just need a guiding document that we can read and then confidently create it.

Looking forward to your valuable response
Re: Pooled country analysis weighting [message #22443 is a reply to message #22441] Fri, 12 March 2021 07:56 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3017
Registered: February 2013
Senior Member
We have had countless postings on the topic of changing the weights and svyset when surveys are pooled. If you search with keywords "svyset" or "pooling", you should find answers to your questions.
Previous Topic: Appending files
Next Topic: Weighting for binary logistic regression analysis in SPSS
Goto Forum:
  


Current Time: Fri Mar 29 04:13:56 Coordinated Universal Time 2024