The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Other countries » Which variable I can use for community level from SADHS
Which variable I can use for community level from SADHS [message #19297] Fri, 22 May 2020 06:37 Go to next message
Frankfeng is currently offline  Frankfeng
Messages: 9
Registered: September 2019
I am planning to study the community effect on one's health outcomes using the South Africa DHS.

I read a paper titled "Individual and community determinants of neonatal mortality in Ghana: a multilevel analysis" that used Ghana DHS.

I am not sure which variable I can use for community level from SADHS? There are v023-stratification or v024-region, but I think it is too large to reflect a community.

I also find that there is v001-cluster in the SADHS, and the DHS webpage says: In most recent DHS surveys, the groupings of households that participated in the survey, known as clusters, are geo referenced. These survey cluster coordinates are collected in the field using GPS receivers, usually during the survey sample listing process. In general, the GPS readings for most clusters are accurate to less than 15 meters.

My question is: Can I use cluster as the community? If not, which one in the SADHS I can use to reflect the community level?

Thank you
Re: Which variable I can use for community level from SADHS [message #19368 is a reply to message #19297] Fri, 05 June 2020 14:51 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2536
Registered: February 2013
Senior Member

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

Yes, users often treat the sample cluster as a community, although this practice has been criticized. It is actually a census enumeration area but it does correspond roughly with a village or a neighborhood. You can construct compositional variables from the individual data within a cluster--for example, the cluster-level contraceptive prevalence rate. (Such estimates have considerable sampling error.) You can also attach geo-coded data. For this survey, I see two spatial data files. ZAGC72DT.dta has "dhscc" as the cluster ID code. ZAGE71FL.dta has "_ID" as the ID code (which matches v001). You can merge the information in these files onto any of the other standard recode files. Let us know if you have difficulty finding those files.
Previous Topic: I can not find the variable G102 for genital cutting in the 2016-17 Haiti dataset
Next Topic: Education of Young Men
Goto Forum:

Current Time: Wed May 18 02:31:41 Coordinated Universal Time 2022