The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Dataset use in Stata » Replicating DHS values
Replicating DHS values [message #6914] Mon, 03 August 2015 04:16 Go to next message
Mzondi Ziba is currently offline  Mzondi Ziba
Messages: 2
Registered: May 2015
Location: Lilongwe
Member
Hello, good people,

My my issue is that, am supposed to replicate the Malawi DHS values on wasting, stunting and underweight for my thesis but am getting different values in stata. for example my sample size for 2010 is 4904 while DHS Malawi used 4849 , and i cant note where the difference is coming from. as such i find, for example stunting to be at 44% not 47% but am also using the same WHO reference and weights, plus removing the flagged values.I also have trouble locating the data of exact date of birth, as am failing to correctly interpret the B3 CMC date. your prompt help will be well appreciated. Regards.I have also attached my do file, so that you help me figure out where i might have messed up.
Re: Replicating DHS values [message #6919 is a reply to message #6914] Mon, 03 August 2015 11:40 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3016
Registered: February 2013
Senior Member
Following is a response from Senior DHS Specialist, Tom Pullum:

I believe the main source of your discrepancy from the report is that you are using the KR file, whereas the DHS report uses the PR file for stunting, underweight, and wasting (as much as possible).

Height and weight of all children under five and all women 15-49 are measured in the household interview. If the mother is alive and in the same household as the child, then the child will also be in the KR file. Otherwise the child will not be in the KR file. The children who are in the PR file but not in the KR file are more likely to be stunted, etc.

In the PR file, the key variables are hc70, hc71, hc72, which match with hw70, hw71, and hw72 in the KR file. The age of the child in months in the PR file is hc1, which is equal to b3 in the KR file.

For your analysis, it should be fine to use just the KR file, for all covariates, not just for those, such as birthweight, that are only in the KR file. This would be better than going back and forth between the PR and KR files. I would just suggest that you include an explanation at the beginning of your report.

Please check whether this is indeed the source of the discrepancy. If not, please let me know. Thanks.


Re: Replicating DHS values [message #6944 is a reply to message #6919] Thu, 06 August 2015 02:36 Go to previous message
Mzondi Ziba is currently offline  Mzondi Ziba
Messages: 2
Registered: May 2015
Location: Lilongwe
Member
Thanks Tom and all. this is very important for me. will come back if i get stuck
again, this should do for now. Regards.
Previous Topic: Institutional deliveries
Next Topic: Calculating median age at first sex
Goto Forum:
  


Current Time: Thu Mar 28 05:37:21 Coordinated Universal Time 2024