The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Merging data files » Merging Malawi MIS 2017 Datasets
Merging Malawi MIS 2017 Datasets [message #16635] Fri, 08 February 2019 00:57 Go to next message
vivianolabamiji is currently offline  vivianolabamiji
Messages: 3
Registered: February 2019
Member
Hello,

I want to look at maternal knowledge of malaria prevention and under-5 parasitemia in the Malawi 2017 MIS but I am unsure of which datasets to merge. The IR file contains questions on malaria prevention, but the KR file does not have the result of the malaria blood smear test. I originally merged the HR and IR files but noticed that there were several variables measuring the same thing. I am now stuck between 1) merging the IR file to the PR or 2) merging the KR to the PR. I just want to link the variables on maternal knowledge of "protection against malaria" to under-5 child blood smear result. Please advise.

Note: I did go through the previous topics, but I am still not sure of which datasets to merge

Best,

Vivian Olabamiji
Re: Merging Malawi MIS 2017 Datasets [message #16696 is a reply to message #16635] Tue, 19 February 2019 18:37 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1671
Registered: February 2013
Senior Member
Following is a response from Lead Malaria Research Analyst, Cameron Taylor:

Thanks for your question. What you need to think through is "what is your unit of analysis". Are you wanting your unit of analysis to be women or children? This impacts the type of merge that you do. In what you describe it sounds like your unit of analysis is the woman, which means your base file is the IR file.

To examine the woman's knowledge in relationship to her child's parasitemia status you will want to merge the PR file into the IR file (each row is information about a woman). The PR file has all the malaria testing information for children under 5. You need to keep in mind that there are women in the IR file who don't have a child under age 5 who was tested for malaria and conversely you have children in the PR file who were tested for malaria but who don't have a mother who was interviewed as part of the women's questionnaire (mother is dead, mother wasn't in the household, the kid was visiting, etc).

If you want the children to be your unit of analysis/denominator you would then merge the PR file into the KR file (each row is information about a child). The malaria knowledge questions are also in the KR file but keep in mind that the unit of analysis in the KR file is children, so this is knowledge of the mother linked to the kid.

Sorry this is very tricky! Be careful about your unit of analysis since that will dictate the type of merge you perform. For more information about DHS data structures and data files please see the following YouTube tutorials

https://www.youtube.com/watch?v=BmiSDPebmgc
https://www.youtube.com/watch?v=fzLNQkkvDeI
Re: Merging Malawi MIS 2017 Datasets [message #16717 is a reply to message #16696] Fri, 22 February 2019 20:40 Go to previous messageGo to next message
vivianolabamiji is currently offline  vivianolabamiji
Messages: 3
Registered: February 2019
Member
Thank you so much for your reply. This information was extremely helpful. My unit of analysis is children, so I went ahead and merged the KR and PR files. Thank you again for your help!

Best,

Vivian Olabamiji
Re: Merging Malawi MIS 2017 Datasets [message #17559 is a reply to message #16696] Mon, 15 April 2019 22:40 Go to previous messageGo to next message
vivianolabamiji is currently offline  vivianolabamiji
Messages: 3
Registered: February 2019
Member
Hello,

I have another question about the country-specific variable s504d. The questionnaire asked women "how can someone protect themselves against malaria?" Some of the responses were "burn leaves, avoid drinking dirty water, avoid eating bad food and avoid getting soaked in the rain" which seem to be traditional practices that are not supported by scientific literature. Did the 2017 Malawi indicator survey classify these practices as false malaria prevention methods? Also, when can I find more information about how this question was designed?

Best,
Re: Merging Malawi MIS 2017 Datasets [message #17612 is a reply to message #17559] Fri, 26 April 2019 11:01 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1671
Registered: February 2013
Senior Member
Following is a response from DHS Lead Malaria Research Analyst, Cameron Taylor:

Thank you for your question. Variable s504d is a country specific variable (note the s in front of the variable name). This variable aligns with Q504 in the woman's questionnaire. As you can see in the dataset each variable aligns with the responses in the questionnaire.
s504a= protection against malaria: sleep under a treated net
s504b= protection against malaria: sleep under an insecticide-treated mosquito net
s504c= protection against malaria: use mosquito net
etc

For this question the interviewer asks "how can someone protect themselves against malaria?" and they categorize the responses given using the answer options below. Note that a woman could mention multiple ways to protect themselves. We include both correct and non-correct ways that someone could protect themselves as answer options because a woman might state a non-correct option and you need to be able to record that option. Also please note that the for social behavior change knowing how many women are stating incorrect answers is just as informative as someone stating the correct way to protect themselves against malaria. As you can see in Table 5.1 in the final report we did not include these incorrect measures in the tabulations for malaria prevention.

/index.php?t=getfile&id=1418&private=0

  • Attachment: Malaria_Q.png
    (Size: 43.13KB, Downloaded 137 times)
Previous Topic: Merging datasets for different countries on SPSS
Next Topic: Merge BR to PR
Goto Forum:
  


Current Time: Mon Oct 14 02:38:59 Eastern Daylight Time 2019