The DHS Program User Forum: Geographic Data » District Codes in recode files

Home » Data » Geographic Data » District Codes in recode files

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

District Codes in recode files [message #122]

Mon, 18 March 2013 14:30

DHS user
Messages: 111
Registered: February 2013

Senior Member

Why are there no labels for the district or municipal variables within the recode files?

Report message to a moderator

Re: District Codes in recode files [message #123 is a reply to message #122]

Mon, 18 March 2013 14:35

ClaraB-DHS
Messages: 38
Registered: March 2013
Location: Maryland

Member

Due to the confidentiality requirements of the data, administrative classifications below the DHS regions level are not able to be included in the recode files. If your analysis will be conducted at a level that does not align with DHS regions, you may use the GPS data within a Geographic Information System (GIS) to join the displaced cluster locations to other data sources. Visit the Available Datasets page: https://dhsprogram.com/data/available-datasets.cfm to view whether GPS data is available for your survey(s) of interest.

[Updated on: Tue, 20 February 2024 07:48] by Moderator

Report message to a moderator

Re: District Codes in recode files [message #208 is a reply to message #123]

Wed, 27 March 2013 09:13

Thea-DHS
Messages: 28
Registered: March 2013

Member

Caution is advised when reclassifying survey clusters to areas that are smaller than DHS regions, as the survey was not designed to be representative below the DHS regions level, which in most countries is the administrative unit 1 (e.g. Province, State). Some cases such as the Malawi 2011 DHS was sampled to be representative at both the district level and the province level.

You may use the GPS data to reclassify the cluster locations from each survey according to their location. The GPS datafiles contain a shapefile of the cluster locations for use in a GIS software package. QGIS (http://qgis.org/) is a good free software option if you do not have access to paid GIS software products. GADM (http://www.gadm.org/) is a good resource of administrative shapefiles that can be used in the reclassification process.

Report message to a moderator

Re: District Codes in recode files [message #690 is a reply to message #208]

Sat, 17 August 2013 20:15

amira.elshal.1@city.ac.uk is currently offline

amira.elshal.1@city.ac.uk
Messages: 52
Registered: April 2013

Senior Member

Dear Sirs,

What about Egypt?

In the results of Egypt's ever-married questionnaire of DHS 2008, how can I know to which district an interviewed woman belong? I can check the governorate, but I cannot find the district. I need to compare between reformed and non-reformed districts.

Thank you.

Sincerely,
Amira

Report message to a moderator

Re: District Codes in recode files [message #702 is a reply to message #690]

Wed, 21 August 2013 14:19

ClaraB-DHS
Messages: 38
Registered: March 2013
Location: Maryland

Member

The same principle applies to the Egypt data as to all other DHS countries.
The recode file does not contain information on the district in which each cluster is located, only the DHS region (which, in the case of the Egypt 2008 DHS correspond to the Urban governorates, the Frontier governorates, Upper Egypt and Lower Egypt). You could import the GPS data points into a Geographic Information System (GIS) software such as ArcMap or QGIS and spatially join the displaced cluster locations to GIS polygon data of the district boundaries (as described in the posts above).

However, as mentioned previously on this forum thread, we do advise caution when conducting your analysis at smaller administrative levels, as the surveys were not designed to be representative below the DHS regions level. Also, please bear in mind that, because the displacement for this survey was restricted to the DHS region level, it is possible that some clusters were displaced across lower level (e.g. district) boundaries during this process. For more information on the displacement process, please see our website http://www.measuredhs.com/What-We-Do/GPS-Data-Collection.cfm .

Report message to a moderator

Re: District Codes in recode files [message #1218 is a reply to message #208]

Thu, 23 January 2014 16:13

yashin5
Messages: 1
Registered: January 2014
Location: Washington DC

Member

Hi,

I was very interested to hear that in some cases, the DHS data may be representative at the district level. I am interested in looking at baseline HIV testing rates (and potentially other indicators) for a district of Nampula Province (Nacala a Velha) in Mozambique and wondered if the most recent DHS or AIS in Mozambique is representative at the district level.

Thanks for your help.

Yashin

ysl

Report message to a moderator

Re: District Codes in recode files [message #1226 is a reply to message #1218]

Fri, 24 January 2014 10:51

Josh-DHS
Messages: 19
Registered: November 2013

Member

Hello Yashin,

It is true that some DHS surveys are designed to give representative results at two levels. This is not the case for any of the Mozambique DHS or AIS surveys which are all representative at the level of the province (1st administrative level). However, if you have a polygon shapefile of the district you are looking at, you can request the GPS datasets for the 2011 DHS and 2009 AIS and do a spatial join as described above to find out which, if any, clusters were located in Nacala a Velha. Do bear in mind though that the samples were not designed to give representative results at that level.

The displacement of the cluster coordinates was restricted to the DHS region (province) for the Mozambique 2009 AIS and to the district (admin2) for the 2011 DHS.

In general, a good way to find out the level at which a given survey's results are representative is to look it up in the boundaries page of the Spatial Data Repository http://spatialdata.measuredhs.com/boundaries.html. You may also download a GIS shapefile of the boundaries from the Spatial Data Repository.

Hope this helps

Report message to a moderator

Re: District Codes in recode files [message #1521 is a reply to message #702]

Thu, 06 March 2014 17:39

amira.elshal.1@city.ac.uk
Messages: 52
Registered: April 2013

Senior Member

Hello Josh,

Is Egypt DHS survey designed to give representative results at the first and second (district) administrative levels. Are Egypt DHS surveys representative at both levels?

Thank you.

Amira

Report message to a moderator

Re: District Codes in recode files [message #1522 is a reply to message #1226]

Thu, 06 March 2014 17:43

amira.elshal.1@city.ac.uk
Messages: 52
Registered: April 2013

Senior Member

Report message to a moderator

Re: District Codes in recode files [message #1531 is a reply to message #1522]

Mon, 10 March 2014 12:55

Josh-DHS
Messages: 19
Registered: November 2013

Member

Hello Amira,

Thanks for your post. Several of the more recent Egypt DHS surveys have samples that are designed to be representative at two levels (specifically, the 2000, 2005 and 2008 surveys). However, these do not exactly correspond to the first and second administrative levels (governorates and regions respectively). At the first level of representativeness, the four DHS regions are groupings of governorates. At the second level, the DHS regions correspond to the individual governorates (with the exception of the 5 frontier governorates which are grouped at this level as well.

A good resource for viewing the DHS region boundaries and comparing them across levels and surveys is the Spatial Data Repository http://spatialdata.dhsprogram.com/boundaries.html?country=EG . There, you can also download the boundaries as GIS shapefiles.

If this has not answered your question adequately, please reply.

Josh

Report message to a moderator

Re: District Codes in recode files [message #2494 is a reply to message #122]

Mon, 30 June 2014 10:00

Philip
Messages: 2
Registered: June 2014
Location: Kenya

Member

Specifically to Malawi, I did a join of DHS shapefile and the Malawi Admin shapefile that has district name and a few of the district names are in disagreement(with district name in the 2010 DHS shapefile). Is it possible that during displacement of GPS some of the data points within a district are displaced to a totally different district other than original district where the survey was conducted?

Report message to a moderator

Re: District Codes in recode files [message #2495 is a reply to message #122]

Mon, 30 June 2014 10:02

Philip
Messages: 2
Registered: June 2014
Location: Kenya

Member

Report message to a moderator

Re: District Codes in recode files [message #2512 is a reply to message #2495]

Wed, 02 July 2014 10:52

Thea-DHS
Messages: 28
Registered: March 2013

Member

Phillip,

Thanks for your post. Please can you confirm that you are using the GPS data file mwge62fl.zip from the Malawi 2010 DHS and the regions boundary shapefile for the same survey downloaded from the Spatial Data Repository ( http://spatialdata.dhsprogram.com/boundaries.html?country=MW)?

The displacement of the GPS data for the Malawi 2010 DHS was restricted at the second administrative level which means that it was ensured that no coordinates were displaced across district boundaries. Therefore, all points in the GPS dataset fall in the correct district and survey region when mapped onto the DHS regions shapefile. Please note that shapefiles from different sources vary according to their level of detail and date created. Therefore, a spatial join using a different shapefile can produce slightly different results for areas that are close to boundaries.

Note that the Malawi 2010 DHS is representative at 2 levels, so the boundaries shapefile downloaded from the Spatial Data Repository includes the boundaries of both the first level (the three administrative regions) and the second level (the 27 districts or groupings of districts). Before conducting the spatial join, you will need to create a separate shapefile of just the level 2 regions by selecting and exporting them.

Also note that there were 22 clusters in that survey that had missing coordinates and so were assigned zero as their latitude and longitude.

If this has not answered your question adequately, please reply.

Report message to a moderator

Re: District Codes in recode files [message #3063 is a reply to message #2512]

Fri, 10 October 2014 02:40

gregounours
Messages: 1
Registered: October 2014
Location: Malawi

Member

Hi,
The link you provided does't seem to be working.
Greg

Report message to a moderator

Re: District Codes in recode files [message #3064 is a reply to message #3063]

Fri, 10 October 2014 09:06

Thea-DHS
Messages: 28
Registered: March 2013

Member

Greg,
The Spatial Data Repository website was recently updated to include the ability to compare subnational boundaries from DHS Program surveys. Please use this updated link to view the boundaries for Malawi: http://spatialdata.dhsprogram.com/boundaries/#view=map&c ountryId=MW&surveyId=333&level=1

You may also download the region boundaries by closing the map window and choosing the survey of interest from the table: http://spatialdata.dhsprogram.com/boundaries/#view=table& ;countryId=MW

Please respond if you have any other questions.

Report message to a moderator

Re: District Codes in recode files [message #10715 is a reply to message #123]

Mon, 05 September 2016 11:33

CKAllen
Messages: 7
Registered: April 2016
Location: London, UK

Member

Is the only way to see the districts to merge with a GPS file (Malawi 2004 and Malawi 2010)? I am wondering if this is the case even for countries that were designed to be representative at the district level (specifically Malawi 2010).

Report message to a moderator

Re: District Codes in recode files [message #18284 is a reply to message #2512]

Mon, 28 October 2019 13:05

zkarar
Messages: 3
Registered: August 2019

Member

DHS webpage on methodology for GPS data (https://dhsprogram.com/What-We-Do/GPS-Data-Collection.cfm) states that "In surveys released since 2009 the displacement is restricted to the country's second administrative level where possible" - can you kindly confirm if it is true for Bangladesh DHS 2014 (i.e. no coordinates were displaced across district boundaries)?
Thanks and Regards,
Karar

Report message to a moderator

Previous Topic:	Linking multiple DHS region levels to survey data
Next Topic:	GPS data missing for Haiti DHS 2012

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Thu May 15 08:44:45 Coordinated Universal Time 2025