The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Weight Data HIV Male Female Merge (Weight Data HIV Male Female Merge)
Weight Data HIV Male Female Merge [message #18858] Mon, 02 March 2020 10:08 Go to next message
behayes4 is currently offline  behayes4
Messages: 13
Registered: February 2020
Member
I combined IR, MR and AR data across multiple nations. From what I can gather across the message boards, this merged data should be weighted only by the AR weight for men. Does this mean I only include HIV05 weight for the male respondents?
Re: Weight Data HIV Male Female Merge [message #18948 is a reply to message #18858] Tue, 24 March 2020 14:22 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1770
Registered: February 2013
Senior Member


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

When doing a merge, I generally retain the weights from the different files. If you are doing analysis that includes the HIV test result, i.e. hiv03, then you use hiv05 as the weight. Otherwise, to analyze other data on men, use mv005, and other data on women, use v005.
Re: Weight Data HIV Male Female Merge [message #18951 is a reply to message #18948] Tue, 24 March 2020 14:29 Go to previous messageGo to next message
behayes4 is currently offline  behayes4
Messages: 13
Registered: February 2020
Member
Thanks Tom!! When using svyset command - it calls for PSU and strata. Do you pull the strata variable from the IR and MR files? Should the HIV cluster data be used for the PSU?
Re: Weight Data HIV Male Female Merge [message #18979 is a reply to message #18951] Mon, 30 March 2020 15:21 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1770
Registered: February 2013
Senior Member

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

After you have done the merge , v023 and hv023 should be identical. Use either of them for the strata. Similarly, the cluster variable should be the same on each merged record--hv001, hv021, v001, hivclust. Except that since you would have used the cluster variable as part of the merge, you may have done some renaming.
Previous Topic: Weighting and Multilevel Logistic Regression
Next Topic: Weight about pooling 2003,2008,2013,2018 BR dataset in Nigeria
Goto Forum:
  


Current Time: Fri Apr 10 01:55:12 Eastern Daylight Time 2020