The DHS Program User Forum      
Discussions regarding The DHS Program data and results
Home » Data » Dataset use (other programs) » Harmonised region variable
Harmonised region variable [message #14010] Fri, 02 February 2018 10:38 Go to next message
lillo?S is currently offline  lillo?S
Messages: 17
Registered: December 2015
Member
Hello,
I was wondering if there was an already-established way of harmonising the region variable over time for all countries in the DHS. For example variable v101 for Uganda in 2000/1 is: 1 Central, 2 Eastern, 3 Northern, 4 Western, while in 2006 it is: 1 Central1, 2 Central2, 3 Kampala, 4 East Central, 5 Eastern, 6 North, 7 West Nile, 8 Western, 9 Southwest. If I need to do an analysis at the regional variable, I guess I should change the categorisation of the 2006 variable into 1 Central including Central1-Central2-Kampala, 2 Eastern including East Central-Eastern, 3 Northern including North-West Nile, and 4 Western including Wester-Southwest. Does this make sense? Is there anyone who has already done this harmonisation or should we do it by looking at survey-specific frq files?
Thank you.
Re: Harmonised region variable [message #14012 is a reply to message #14010] Fri, 02 February 2018 12:14 Go to previous messageGo to next message
boyle014 is currently offline  boyle014
Messages: 20
Registered: December 2015
Location: Minneapolis
Member
Hi lillos.

Yes, this kind of harmonized data is available through IPUMS-DHS. You can log in to IPUMS-DHS using your DHS username (email) and password. Select the Ugandan samples you want. You will find the harmonized region variables under Geography -> Integrated Geography. The harmonization was done by professional geographers at the University of Minnesota Population Center.

For Uganda, you have two options. If you want to do regional comparisons all the way back to 1995, there are four comparable regions. If you just want to compare 2005 to 2011, there are nine comparable regions. (This difference occurs because the older surveys tended to identify fewer regions and because boundaries tend to change over time.)

I'm attaching a copy of the IPUMS-DHS User Guide, which walks through all the steps for creating and downloading DHS data using this tool. I think you'll find the process is pretty straightforward.

Best,
Liz Boyle


Professor Elizabeth Boyle
Sociology & Law, University of Minnesota, USA
Principal Investigator, IPUMS-DHS
Re: Harmonised region variable [message #14017 is a reply to message #14012] Sat, 03 February 2018 07:02 Go to previous messageGo to next message
lillo?S is currently offline  lillo?S
Messages: 17
Registered: December 2015
Member
Hi Liz,
Thank you for your reply. Yes, I also considered IPUMS-DHS but I'd need this for all countries for which the DHS is available, while IPUMS-DHS is currently available for African countries and for DHS only so if what you say is feasible for all DHS African countries, I would still need to do it for non-DHS (MIS, AIS) and for all DHS (including MIS and AIS) for non-African countries. Please correct me if I'm wrong. Thanks.
Re: Harmonised region variable [message #14070 is a reply to message #14017] Fri, 09 February 2018 09:19 Go to previous message
boyle014 is currently offline  boyle014
Messages: 20
Registered: December 2015
Location: Minneapolis
Member
Dear lillos,

Yes, that's true. IPUMS-DHS currently includes the standard DHS for African and some south Asian countries. We are adding more surveys as quickly as we can, but we have limited funding and thus limited staff. We do appreciate every dollar we get from NICHD though!

Sorry to hear that IPUMS-DHS can't solve your issue. I don't know of anyone who has already harmonized all the geography across all DHS samples.

Yours,
Liz


Professor Elizabeth Boyle
Sociology & Law, University of Minnesota, USA
Principal Investigator, IPUMS-DHS
Previous Topic: import data file to R
Goto Forum:
  


Current Time: Sat Feb 17 22:09:29 Eastern Standard Time 2018