Using a sub-sample for analysis [message #10893] |
Thu, 29 September 2016 21:48 |
Ahmed M Kamel
Messages: 5 Registered: September 2016 Location: Egypt
|
Member |
|
|
Thanks alot for your efforts.
I wanted to know if it's possible to extract a random sample from the nearly 14000 cases in the DHS survey in order to perform statistical analysis instead of conducting the analysis on allc cases
|
|
|
Re: Using a sub-sample for analysis [message #10907 is a reply to message #10893] |
Tue, 04 October 2016 17:05 |
admin
Messages: 50 Registered: November 2012
|
Senior Member Administrator |
|
|
Dear User, Here is a response from one of our technical experts, Dr. Lia Florey:
Quote:Hello,
For the analysis you propose you will need to use the person recode file "CMPR60FL.sav". The data on individual net use and on malaria status of children will be available in that data file. In order to limit the data to the appropriate age group you can use the variable hml16a which gives age in months. You will want to limit your analysis to children between the ages of 6-59 months as that is the age range of children who were eligible for malaria testing:
SELECT IF (hml16a>5 AND hml16a<60).
Other useful variables are listed below:
Hml33==0 "malaria measurement" (you will want to limit your analysis to those children who had a valid malaria test result available).
Sh418==1 "Pf positive"
Sh418==2 "Other Plasmodium positive"
Sh418==3 "positive mixed species"
You will probably want to create a y/n variable for any positive test which would include values of 1, 2, or 3 for the variable sh418.
Hml10 ==1 "person slept under an ITN the previous night"
Hml20==1 "person slept under an LLIN the previous night"
Hml12==1,2 or 3 "person slept under any net the previous night"
In this dataset you can also find information about the sex, age, haemoglobin level, household wealth quintile, mother's education, region, urban/rural residence, etc. for the children. Any data on child health gathered from the women's questionnaire will not be in this datafile but rather in the KR datafile and would need to be merged into this datafile to be useful.
Please note the disclaimer that I am a Stata user and not an SPSS user so please adjust the above accordingly.
Best
We now have repositories of code written in Stata and SPSS available on Github. Please reference these code repositories as a resource for code for matching or calculating DHS indicators. The code repositories can be found at:
https://github.com/DHSProgram/DHS-Indicators-Stata
https://github.com/DHSProgram/DHS-Indicators-SPSS
|
|
|
|