The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Service Provision Assessment (SPA) » Senegal 2018 and 2019 ftype (urban / rural classification)
Senegal 2018 and 2019 ftype (urban / rural classification) [message #23519] Wed, 29 September 2021 22:14 Go to next message
hannah.leslie is currently offline  hannah.leslie
Messages: 11
Registered: September 2015
Member
Hello DHS experts,

I'm puzzled by the variable ftype - urban / rural - in the 2018 and 2019 facility datasets for the Senegal continuous SPA. The percent of urban facilities increases dramatically from prior surveys, and all (2019) or nearly all (2018) all non-health huts are classified as urban in these two datasets, even in regions that previously had almost no urban facilities (Kedougou, Matam, Sediou). There's no indication that the sampling changed from 2017, and the other proportions seem consistent with prior years. The sampling description in the guides still indicates that health huts are sampled based on their attached health post, so it's quite odd that nearly all health huts are rural but the health posts are almost 100% urban. In addition, many of the facilities listed as urban in these surveys were classified as rural in 2016 and 2017 based on linking them using location or IDs from ANDS.

Was there a change in the definition of urban between 2017 and 2018 or might there be an error in this variable?

Thank you!

Hannah
Re: Senegal 2018 and 2019 ftype (urban / rural classification) [message #23524 is a reply to message #23519] Thu, 30 September 2021 11:01 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3208
Registered: February 2013
Senior Member

Following is a response from DHS Lead Statistician, Ruilin Ren:

Thanks for your email and the question raised.

Regarding the urban-rural variable in SPA data, firstly, it is not a survey control variable because it does not exist in the health facility master list. Urban-rural is collected in the field based on the location of the facility and the urban-rural classification of the country.
Regarding the Senegal case specifically, firstly, Senegal's urban-rural classification was modified in 2017, this may affect the comparability of the urban-rural in later surveys.

Secondly, for the health huts, though they are attached to health posts in management, most of them are located in rural areas and remote areas. Their urban-rural definition can be different from that of the health post they are attached to.

Re: Senegal 2018 and 2019 ftype (urban / rural classification) [message #23525 is a reply to message #23524] Thu, 30 September 2021 12:25 Go to previous messageGo to next message
hannah.leslie is currently offline  hannah.leslie
Messages: 11
Registered: September 2015
Member
Thank you Bridgette and Ruilin for the quick response. Yes, I understand urban/rural classification is not used in the sampling strategy for SPA.

That is very helpful to know that the classification changed between 2017 and 2018. Is there anywhere that describes this change? It's not mentioned in the 2018 reports for DHS or SPA; the 2018 SPA report states that 55% of the population lives in rural zones. I don't see any major change in % population rural in international databases. If there is any further detail available to help clarify the source or content of the change, I would greatly appreciate it. I will try to follow up with ANSD as well.

Many thanks,

Hannah
Re: Senegal 2018 and 2019 ftype (urban / rural classification) [message #23526 is a reply to message #23525] Thu, 30 September 2021 14:58 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3208
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

Hi Hannah--Apparently we have nothing to add on this topic, and there's no alternative to checking with ANSD. We just know that there was a change in the sampling frame and a change in the definition of urban/rural around 2017. Hope you can find more documentation and I'm sorry that we don't have records on this. Best regards--Tom

[Updated on: Thu, 30 September 2021 14:59]

Report message to a moderator

Re: Senegal 2018 and 2019 ftype (urban / rural classification) [message #23527 is a reply to message #23526] Thu, 30 September 2021 15:20 Go to previous message
hannah.leslie is currently offline  hannah.leslie
Messages: 11
Registered: September 2015
Member
Thanks Tom! I appreciate the clarification. If I find out anything through other avenues, I'll follow up here.

Best,

Hannah
Previous Topic: SPA statistics
Next Topic: About one SPA variable
Goto Forum:
  


Current Time: Thu Dec 12 08:16:56 Coordinated Universal Time 2024