The DHS Program User Forum: Geographic Data » A general approach to merge GPS and Individual Data

Home » Data » Geographic Data » A general approach to merge GPS and Individual Data

Show: Today's Messages :: Show Polls :: Message Navigator

A general approach to merge GPS and Individual Data [message #9316]

Wed, 09 March 2016 13:02

Doon
Messages: 1
Registered: March 2016

Member

I am doing an analysis that requires geographic information. I downloaded the GPS, Individual Recode, and Men Recode for all sub-Saharan African countries when available. I have a file of all individual recodes for different countries and phases, all appended and a separate shapefile with all the available GPS files merged. I have a question regarding the approach I need to use to merge the geographic information to the individual files.

In the geographic files, the identifier is of the form:

I use the case of Liberia, phase 5 as an example (File LBGE5CFL.shp)

Survey = Two letter country code + Year , ex: LB2009
Cluster = Cluster number, ex: 1
=> I thought a unique identifier would be of the form LB2009_1

However, it is not easy to match such information with the individual recode file.
There are 2 Stata files that can be downloaded: (LBIR51FL and LBIR5AFL)They both have the country code LB5, although they run for years 2007 to 2009. My question is then: are they run in the same villages, meaning that I can merge all the clusters in these files with those in LBGE5CFL.shp, or can I just only merge the subsample of the individual recode that were collected in 2009?

In other words,
What is the best way for me to create a unique identifier in the appended file of *all* the individual recodes in sub-Saharan Africa that I'll manage to merge correctly with the GPS files, provided that phase of the DHS collected GPS information?

Report message to a moderator