The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » HIV » HIV testing response rate
Re: HIV testing response rate [message #18826 is a reply to message #18825] Wed, 26 February 2020 14:39 Go to previous messageGo to previous message
behayes4 is currently offline  behayes4
Messages: 24
Registered: February 2020
Member
As an example the Burundi 2016/2017 HIV dataset has 21,955 cases. I merged the HIV data into the individual respondent data (which resulted in 8,527 valid cases) and the male respondent dataset (which resulted in 7,377 valid cases). This is only 15,904 observations across the two datasets when there theoretically should be 21,955 - a difference of 6,051 cases. Burundi 2016/17 has one of the biggest dropoffs. I did this for 30 other datasets. The majority have a few hundred that drop when merged. Can you explain why each case in the HIV dataset does not merge into a case for either the IR or MR datasets? Where do the remaining respondents merge into?
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: Difference in N when HIV data merged
Next Topic: Merging Women-Men-HIV Data: Different Countries, Different Years
Goto Forum:
  


Current Time: Thu Dec 12 00:35:38 Coordinated Universal Time 2024