| Re: Children age Namibia DHS VI [message #19699 is a reply to message #19682] |
Tue, 04 August 2020 13:11  |
Bridgette-DHS
Messages: 3040
Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Research & Data Analysis Director, Tom Pullum:
Please check the description of this survey in the main report. When any survey has large numbers of Not Applicable (NA) cases, indicated in Stata with a period, it is usually because the survey involved subsampling. It appears that the children in only about a third of the households were selected for hemoglobin testing. Those are the only children who have a valid (not NA) value of hw1.
Note that the KR file always includes all children born in the past five years. Some of them had died (b5=0) and were not tested even though their household may have been selected for the subsample.
The malaria variables are in the KR file. No merging is necessary to get, for example, ml101. But you need to distinguish between malaria variables that refer to the child and others that refer to the mother, whose data are on the child's record. I recommend that you look at the questionnaire, which is in an appendix to the main report.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
