| Trend analysis for Nutritional status of children [message #16951] |
Thu, 14 March 2019 09:45  |
Sasmita
Messages: 3
Registered: February 2019
Location: China
|
Member |
|
|
Dear DHS experts,
I am doing a trend analysis of Nepal DHS data on the nutrition status of under 5 years in different waves from 2001-2016. I merged PR and KR file as I am interested to see both children and maternal-related variables. I am using SPSS.
Problem 1
While working for the data analysis, I could replicate the exact figure from the report for stunting only for the year 2016 data. And could not find a similar figure as in report for other waves (2001, 2011, 2016) with similar consideration. for example, I get 2485 instead of 2475 in 2011 and 5258 instead of 5262 in 2006. I have tried many times and I am taking into consideration:-
a) sample weight variable (HV005/1000000)
b) selected de facto children using HV103
Syntax used for Stunting.
recode hc70 (lo thru -201 = 1) (-200 thru 9990 = 0 ) (else=sysmis) into stunting.
variable labels stunting "Stunting".
value labels stunting 0 "Not stunted" 1 "Stunted".
formats stunting (f1.0).
compute wgt = hv005/1000000.
weight by wgt.
compute filter_$=(hc70 < 9990 & hv103=1).
variable labels filter_$ 'hc70 < 9990 & hv103=1 (FILTER)'.
value labels filter_$ 0 'Not Selected' 1 'Selected'.
formats filter_$ (f1.0).
filter by filter_$.
execute.
frequencies stunting.
Problem 2 :
a)Should I merge the HW file to PR file then only merge the combined file to KR file? Or merge HW file with merged datasets of KR and PR file?
I would be grateful if you can share the ideas for solving these issues. I am struggling with this difference in figures.
Thank you.
[Updated on: Thu, 14 March 2019 09:54] Report message to a moderator |
|
|
|
| Re: Trend analysis for Nutritional status of children [message #19196 is a reply to message #16951] |
Fri, 08 May 2020 12:32  |
Trevor-DHS
Messages: 788
Registered: January 2013
|
Senior Member |
|
|
1) My supposition is that in the process of finalizing the survey datasets there have been some data updates that have affected a few cases. I am matching the same number as you for the 2011 survey (2485). This sometimes happens in the process of finalizing datasets, where the reports may be based on earlier versions of the datasets. The percentage stunted, though, should be virtually identical to the report - I get 40.5% stunted.
2) For the HW files, a few steps:
a) Check if hc70-hc72 exist in the PR file. If they do, then you don't need the HW file as the data are already in the PR dataset.
b) Check if hc1 exists in the PR file. If it does exist then the anthropometry data was collected at the household level, andn you should merge the HW file to the PR file.
c) If the hc variables do not exist in the PR file, but the HW variables in the KR contain data (see hw1, hw2, hw3 for example) then the data were collected in the women's questionnaire and only exist for children of women interviewed. In this case, merge the HW file to the KR file. You probably do not want to analyze the data with the PR file in this case as it was limited just to children of interviewed women and there will be many children in the PR file who do not have data as their mother was not interviewed.
|
|
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
