The DHS Program Geospatial FAQ

23/04/2020

Displacement

Q1: Are the GPS points in the actual locations that the survey took place in?

A1: No, the points are moved slightly through a process called displacement/jittering. For more information, please see this report.

Access

Q2: How to I access/request the GPS data?

A2: Please see this write-up on our user forum for an in-depth guide to requesting the GPS.

Missing Points

Q3: Why are some points at (0, 0)?

A3: Points we couldn't verify the location of are marked as MIS in the SOURCE column and placed at (0, 0)

Joining GPS to the Recodes

Q4: How do I join the GPS data to the recode data?

A4: Here is some STATA code to do the join. The basic idea is that DHSCLUST and v001 are the join fields. One thing to remember is that the GPS is at the cluster level while recodes are lower.

Undisplaced Data

Q5: Can I please get the interview locations/undisplaced data?

A5: No, the actual locations cannot be released to outside researchers. Even at The DHS Program, access to the undisplaced data is limited to a small number of use cases by our IRB.

District Codes

Q6: Can I get the names of [low-level admin units]?

A6: Our IRB restricts us from releasing non-GPS identifiers below the Admin 2-level (districts in most DHS countries)

Admin 3 Names

Q7: In the recode, there are district codes without names. Can I get them?

A7: We currently do not maintain code books for the district codes. For sub-district codes, please see A6

Point Reuse

Q8: Can I use the 2018 points with the 2011 survey?

A8: Samples are independently generated for each survey. It is highly unlikely that two rounds of the DHS would visit exactly the name location.

Other Questions

Q9: I have another question not answered in this FAQ

A9: Please email it here and we will try to get back to you with a response: gpsrequests@dhsprogram.com. Before you email us, please read all of the documentation provided with the GPS Data