The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » HIV » Merging Female (IR) file and Male (MR) files (Merging Female and male file to generate combined values for variables )
Merging Female (IR) file and Male (MR) files [message #27253] Fri, 07 July 2023 06:44 Go to next message
G Kumar S is currently offline  G Kumar S
Messages: 2
Registered: July 2023
Member
Hello all,
I am in the in the process of analyzing the Indian DHS data (NFHS-V (2019-21)). While looking for some references, I came across a paper titled 'Knowledge of HIV/AIDS and its determinants in India: Findings from the National Family Health Survey-5 (2019-21)'.

This paper merged the HIV module (respondents aged 15-49 years) female (108785) and male (93267) files of NFHS-5 (2019-21) into a single dataset. The outcome variable of the paper was Knowledge of HIV/AIDS, with gender as one of the independent variables of the analysis.

1. Is this okay to combine these files for deriving combined figures?
2. If so, what weights are to be used?
Thank you in advance.

The link of the paper is
http://www.populationmedicine.eu/Knowledge-of-HIV-AIDS-and-i ts-determinants-in-India-nFindings-from-the-National-Family, 163113,0,2.html
Re: Merging Female (IR) file and Male (MR) files [message #27269 is a reply to message #27253] Mon, 10 July 2023 10:46 Go to previous messageGo to next message
Janet-DHS is currently offline  Janet-DHS
Messages: 888
Registered: April 2022
Senior Member
Following is a response from DHS staff member, Tom Pullum:

Men were subsampled in NFHS-4 and -5, so there are only about 1/6 as many men as women. If the cases for women and men were put into a single file, and a weight was calculated as v005 for women and 6*mv005 for men, the weighted sample would be approximately representative of the population of women and men combined. This simple calculation of a combined weight is probably not optimal but would be useful. (In regressions these weights would automatically be re-normalized to have a mean of 1.) It would be safest to give separate estimates for women and men separately rather than for the pooled sample.
Re: Merging Female (IR) file and Male (MR) files [message #27275 is a reply to message #27269] Wed, 12 July 2023 00:29 Go to previous message
G Kumar S is currently offline  G Kumar S
Messages: 2
Registered: July 2023
Member
Thank you very much for the clarification.
Previous Topic: HIV Testing - Nigeria
Next Topic: Modern Contraceptive Use for women with higher risk of HIV
Goto Forum:
  


Current Time: Mon Nov 25 23:18:24 Coordinated Universal Time 2024