| Re: Comparison - Continuous DHS Senegal [message #19959 is a reply to message #19879] |
Tue, 08 September 2020 09:16   |
Bridgette-DHS
Messages: 3032
Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Lead Statistician, Ruilin Ren:
Regarding the sampling frame used for the Senegal continuous DHS, the 2013-2014 surveys were based on 2002 census, the 2015-2016 surveys were based on 2013 census, while the 2017 was based on 2013 census but with revised urban-rural. The different versions of the sampling frame differ mainly on urban-rural difference, there were no changes on the definition of regions. So if your comparison is aimed at regional level, there is no concern. But the urban-rural level comparison may have less meaning because of the change in the urban-rural definition.
The surveys are not all independent. The 2013 and 2014 samples are associated, the 2015 and 2016 samples are associated. The 2017 survey was standalone. We have published combined data for the 2013-2014 and 2015-2016, it is better that you use the combined data because the weight variables are already properly calculated in the combined data. However, when you combine the 2013-2014 with 2015-2016 and 2017, you need to pay attention to properly de-normalize the weight.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
