The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Geographic Data » Pakistan DHS 2012/13: Any geographic information?
Re: Pakistan DHS 2012/13: Any geographic information? [message #3747 is a reply to message #3737] Wed, 04 February 2015 16:52 Go to previous message
ClaraB-DHS is currently offline  ClaraB-DHS
Messages: 38
Registered: March 2013
Location: Maryland
Member
The District classification for the Pakistan 2012 survey is attached to this post.

Household selection within a cluster is discussed elsewhere such as here (among others): http://userforum.dhsprogram.com/index.php?t=tree&th=982& amp;goto=1710&S=a62297a3e9176c3b004c51934fd4b1bb#msg_171 0

However, the basic summary is that the household selection within a cluster is random.
The issue regarding the below-region level analysis has to do with the cluster selection not the household selection. Clusters are selected using an inverse proportional to size method. That process is based on the population size of EA and not geographical location or geographic size of an EA. It is possible to have districts or areas within a region that have no selected DHS clusters, there is likely people living in that place they just didn't happen to be selected for the survey.
This causes issues when trying to use the cluster level estimates for interpolation or district level estimates. More on issues of interpolation can be found in Spatial Analysis Report 9, http://dhsprogram.com/publications/publication-SAR9-Spatial- Analysis-Reports.cfm

 
Read Message
Read Message
Previous Topic: Merging Ghana 2003 GPS and Survey data
Next Topic: 44 clusters outlier in DRC 2014 DHS GIS data
Goto Forum:
  


Current Time: Mon Jul 22 08:50:38 Eastern Daylight Time 2019