The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Nutrition and Anthropometry » incorrect stunting rates
incorrect stunting rates [message #14227] Wed, 07 March 2018 08:45 Go to previous message
margovg is currently offline  margovg
Messages: 3
Registered: March 2018
Location: Amsterdam
Member
Hello,

Im trying to calculate the prevalence of stunting using the Burundi 2010 DHS dataset.
I have used the PR file and - presumably- the correct survey settings, the overall percentage that I have calculated is correct (57,7% stunted), however, when I try to calculate the sex specific prevalence I'm of by a bit.



I am using R:

# making a variable for stunting
data2$stunt <- ifelse(data2$hc70 < -200,'stunted','not_stunted')

#survey settings

int.design <- function (){
data.w <<-
svydesign(
id = ~ hv001 ,
data = data ,
weight = ~ hv005 ,
strata = ~ hv022)
}

int.design()

> svyby(~stunt, ~hv104, data.w, svymean)
hv104 stuntnot_stunted stuntstunted se.stuntnot_stunted se.stuntstunted
male male 0.3750139 0.6249861 0.01372043 0.01372043
female female 0.4729965 0.5270035 0.01436399 0.01436399


According to the DHS report 62,1% of boys is stunted and 53,1% of the girls. As you can see from the output above, my percentages come down to 52,7% for girls and 62,5% for boys.

I have deleted from my dataset:
- Children with missing data for variable HC70
- Children with HC70>9000
- Children who did not sleep in the hh the night before the survey(hv103==1)

Can anybody tell me what I doing wrong, and how I can get the correct percentages?

Kind regards,
Margo



[Updated on: Wed, 07 March 2018 08:45]

Report message to a moderator

 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: DHS of India 2015/16 and 2005/06
Next Topic: Consent for additional tests
Goto Forum:
  


Current Time: Thu Mar 28 06:09:15 Coordinated Universal Time 2024