The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Domestic Violence » Differences in observations in Original Stata file and Report
Differences in observations in Original Stata file and Report [message #26900] Mon, 22 May 2023 02:19 Go to next message
uuupaudel22@gmail.com is currently offline  uuupaudel22@gmail.com
Messages: 2
Registered: May 2023
Member
Question: I am working in Nepal 2016 NDHS data set - Domestic violence module

Total number of ever-married women is 3562 in report, but variable to represent ever married women sample not found in dataset. And sample for DV module is 4444 in report but the of original dataset shows total number of observation in physical violence variables (for eg d105a) have 3826 observations. How can I match this 4444 and 3826 obs?

I hope your kind suggestions.

[Updated on: Mon, 22 May 2023 02:37]

Report message to a moderator

Re: Differences in observations in Original Stata file and Report [message #26914 is a reply to message #26900] Wed, 24 May 2023 13:59 Go to previous messageGo to next message
Janet-DHS is currently offline  Janet-DHS
Messages: 657
Registered: April 2022
Senior Member
Following is a response from DHS staff member, Tom Pullum:

Please be more specific. DHS works in more than one country whose name begins with the letter N. I think you are referring to the Nepal 2016 survey and the d105* variables. When I open the IR file for this survey and enter "tab d105a" I get a total of 3,826. That number includes all ever-married women who were respondents for the DV module. If you enter "tab d105a v501" you will see the number 3,826 again. How did you get 3,562?
Re: Differences in observations in Original Stata file and Report [message #26916 is a reply to message #26914] Wed, 24 May 2023 23:48 Go to previous messageGo to next message
uuupaudel22@gmail.com is currently offline  uuupaudel22@gmail.com
Messages: 2
Registered: May 2023
Member
Thank you for your reply.
Yes, I also got the same number of observations 3826 in dataset. But the total observations in published report (Nepal NDHS 2016 - table 19.6) for ever-married women is 3562. I confused with these different total observations for same variable. Should I follow the total observations given in dataset or any adjustment is required in the dataset if any?

Please suggest me.

[Updated on: Wed, 24 May 2023 23:49]

Report message to a moderator

Re: Differences in observations in Original Stata file and Report [message #26943 is a reply to message #26916] Tue, 30 May 2023 17:00 Go to previous message
Janet-DHS is currently offline  Janet-DHS
Messages: 657
Registered: April 2022
Senior Member
The numbers in the final reports are weighted.  Below I will give the distribution of d105a, for example, unweighted and then weighted (with d005).  The unweighted total is 3826 and the weighted total is 3562 (rounded to the nearest integer). The table you are referring to is 16.9, not 19.6.

/index.php?t=getfile&id=2085&private=0
Previous Topic: Husband's family history of domestic violence
Next Topic: Weights for Men's outcomes from the DV subsample
Goto Forum:
  


Current Time: Thu Mar 28 08:13:04 Coordinated Universal Time 2024