| Missing values for anemia [message #9188] |
Fri, 19 February 2016 16:36  |
clarapd
Messages: 8
Registered: November 2015
|
Member |
|
|
Hi everyone, I'm new in this forum.
I'm trying to use the prevalence of anemia among women of reproductive age in several Sub-Saharan countries in a regression.
First, I merged the individual dataset with the household dataset with v001 and v002. Now, to match the anemia data (ha57_x, I understant that each one of these variables in linked to a woman in reproductive age of the household) with each woman I'm using the variable v003 (respondent's line number) and the ha0_x (Index to the household schedule):
gen anemia=ha57_1 if ha0_1=v003
replace anemia=ha57_2 if ha0_2=v003 ...
I'm doing something wrong because the number of missing values I get is always higher than in the report of the survey.
Can someone help me?
Many thanks,
Clara
Clara Pons Duran
[Updated on: Mon, 22 February 2016 11:00] Report message to a moderator |
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
