How to interpret HW70 [message #16699] |
Wed, 20 February 2019 15:13 |
dnameispaone
Messages: 21 Registered: February 2019
|
Member |
|
|
Dear All,
I am planning an analysis of NFHS-4 data. I met with an unexpected problem. I don't know how to interpret "HWw70" variable; I want to determine which all children are stunted, severely stunted and who all are normal. I do not understand how to manipulate this variable HW70 to determine whether a child has stunting or not.
[Updated on: Wed, 20 February 2019 15:18] Report message to a moderator
|
|
|
Re: How to interpret HW70 [message #16700 is a reply to message #16699] |
Wed, 20 February 2019 15:26 |
Reduced-For(u)m
Messages: 292 Registered: March 2013
|
Senior Member |
|
|
HW70 is the height-for-age z-score based on WHO standards, although the DHS provides the values times 100 (so HW=100 would be HAZ of 1).
Stunted is defined as more than 2sd below the cutoff, which is -200 in the DHS coding (a "1" on the WHO scale means 1 standard deviation (sd)).
To define stunted children (using Stata syntax).
Gen stunted = hw70<-200
gen severely_stunted = hw70<-300
You may have to trim off the values of HAZ that are above/below +/-600 as that is the "valid" range of values (I can't remember if DHS already cuts them). But that is the general idea - you just have to define an indicator variable that takes the value of 1 for children with HAZ below -200 (-2 in WHO speak) and 0 for children with HAZ above that cutoff.
|
|
|
|