Following is a response from DHS Senior Research Associate, Cameron Taylor:
Hi Nelly,
Here are the answers to your questions. When you are done with your report please share it with us at DHS. It seems like this is part of a much bigger report that is highlighting DHS malaria data.
- Cameroon 2011: here I guess I need to select only households not selected for the men's questionnaire? i.e. hv027=0?
• Please see this previous user forum post. You are correct in that you should restrict to "if hv027==0"
http://userforum.dhsprogram.com/index.php?t=msg&goto=111 05&&srch=cameroon+ITN#msg_11105
- Chad 2014-2015
• Same as Cameroon, you need to restrict to households not selected for the men's questionnaire
tab ITNinHH if hv027==0 [iweight=wt]
- Guyana 2005: I found different results by wealth quintile (I have used hv270). Do I need to use a different variable?
• I am matching the report when using variable hv270. I checked StatCompiler (which I am assuming you were using) and the results are different on StatCompiler. I will check with our StatCompiler administrator to get that changed.
g ITNinHH=0
forvalues x=1/2 {
replace ITNinHH=1 if hml10_`x'==1
}
tab ITNinHH hv270 [iweight=wt], col
• Kenya surveys are slightly different with their ITN variables in that they restrict the indicator to nets that have been treated within 6-months instead of the typical 12 months. You will see country specific variables (sh or s variables) in more recent datasets that include a specific 6-month retreatment. If you want to match the report you should use the following code. If you want the indicator to be the same in a multi-country analysis I would exclude the hml9 variable which is specifying time since last treatment.
gen wt=hv005/1000000
g ITNinHH=0
forvalues x=1/6 {
replace ITNinHH=1 if hml10_`x'==1 & hml9_`x'<6
}
tab ITNinHH [iweight=wt]
• I am matching the report using
g ITNinHH=0
forvalues x=1/2 {
replace ITNinHH=1 if hml10_`x'==1
}
tab ITNinHH [iweight=wt]
- Zambia 2001-2002: I could not find the variable hv270 in this dataset. Have you used another variable than this one?
• Yes this dataset does not include a wealth index variable. You will need to download the wealth index file ZMWI41FL.DTA and merge it into the HR file.
I also found results for Senegal 2015 (for total and all subdivisions) and Mali 2010 (total and by residence only) => Is there a particular reason why those results are not in the stat compiler?
• We are still working on getting Senegal 2015 put on StatCompiler. Since it is a continuous survey it adds a layer of complication. The Mali 2010 A&P survey is a non-standard survey which is why it is not on Stat Compiler.
Please let us know if you have additional questions
Cameron
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
