District Labels in Ghana DHS [message #5577] |
Wed, 10 June 2015 10:26 ![Go to next message Go to next message](/theme/default/images/down.png) |
ksarnoff
Messages: 2 Registered: October 2014
|
Member |
|
|
Hello,
The Ghana 2003 and 2008 datasets have two variables 'sdist' and 'shdist' that correspond to 18 districts. Are these districts based on some class of administrative divisions, or are they constructed by DHS? I know that district names are suppressed for privacy reasons; however, is the district labeled district 1 in 2003 the same as the district labeled 1 in 2008 (and so on)? In addition, since clusters are displaced, is the district an individual is assigned to in the 'sdist' and 'shdist' variables based on their displaced coordinates? Lastly, will 2003 and 2008 data not be representative at the district level?
Thank you very much,
Kim
|
|
|
Re: District Labels in Ghana DHS [message #6664 is a reply to message #5577] |
Tue, 23 June 2015 13:18 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Matt-DHS
Messages: 18 Registered: July 2014
|
Member |
|
|
The variables 'sdist' and 'shdist' are survey-specific and do not necessarily align with any established administrative divisions. Since they are specific to each survey, they might not correspond to the same area over time. The assignment of these variables would come from the sampling frame and not the location of the GPS coordinate, so the location of the displaced coordinate would not affect it. Keep in mind that the displaced coordinates for these surveys are only guaranteed to be located within the proper region, and could have been displaced across district borders. Both the 2003 and 2008 surveys are only representative at the region level, and not the districts.
|
|
|