Replicating DHS report on Child Health [message #19846] |
Thu, 20 August 2020 04:00 |
Mohammed Beshir
Messages: 1 Registered: August 2020
|
Member |
|
|
Hello Everybody!!
I am from Ethiopia and I have some questions.
When I was working on child health data using Kr file following the "DHS statistics Guideline", I get values with absolutely not matching in any way with DHS report for Ethiopia.
Here are the steps I have used:-
1 - Export All data from ETKR51FL.SAV to ETKR51FL.csv
2 - Then, Using DHS Statistics Guideline I have filtered the required columns from CSV and then counted and calculated percentage (i.e Total percentage of alive children with a very small size or weight) but, the outcome wasn't ídentical with the report. for example, I got 19.76, but the real results are 20.5. So where does the difference come from?
3 - There is also some inconsistency when trying to replicate the values for background characteristics which are indicated in the report. But in a few cases I have found the same result with the report.
What shall I do to avoid this inconsistency for all background characteristics(like region, education level, and so on.)
FYI I have attached below, my calculation outputs in CSV file for Results or output background characteristics on Child health.
[Updated on: Thu, 20 August 2020 04:03] Report message to a moderator
|
|
|
Re: Replicating DHS report on Child Health [message #19857 is a reply to message #19846] |
Fri, 21 August 2020 09:40 |
Trevor-DHS
Messages: 805 Registered: January 2013
|
Senior Member |
|
|
It doesn't look like you have used the sample weights. Please read the section of the Guide to DHS Statistics about weights. You can also view the Youtube tutorials about use of weights. See the videos about
Part I: Introduction to DHS Sampling Procedures
Part II: Introduction of Principles of DHS Sampling Weights
You might also be interested in the videos on Matching DHS Final Report Tables, however I would watch the videos about Sampling and weighting first.
|
|
|