The DHS Program User Forum: Kenya » De jure region of residence (v139)

Home » Countries » Kenya » De jure region of residence (v139) - Kenya 2014

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

De jure region of residence (v139) - Kenya 2014 [message #29202]

Thu, 09 May 2024 23:12

dperezme
Messages: 2
Registered: October 2022

Member

Dear all,

I am currently working on a project on the evolution of the nutritional status of children under 5 in Sub-Saharan Africa and its determinants. For this, I am using all available SSA countries from DHS-V to DHS-VIII. For my analysis, I include regional fixed effects, using for that the variable v139, de jure region of residence. However, in the case of Kenya, the following is happening: a) in the DHS-V, the categories of this variable correspond to the regions in which the country is divided, so that it is ok; b) in the DHS-VIII, the same occurs but disaggregated into counties, so it's ok.

However, in the DHS-VII, Kenya 2014, the categories correspond neither to the regions in aggregate form (as in DHS-5) nor in county form (as in DHS-8), but only to a few counties (8 to be precise), several of which belong to 2 or 3 regions only. What is happening in this case? This is the first time I see this in the DHS, and I don't know how to proceed to homogenise the fixed effects between waves (e.g., we can group the DHS-8 counties to get the DHS-5 regions, but I cannot do the same with the small number of counties in DHS-VII).

What is the reason for this "error" in DHS-VII? What would you recommend me to do in this case? Use de facto region v024? Although I know that v024 and v139 have a different nature: v024 is where the mother was interviewed and v139 is where she actually lives.

Thank you very much for your attention and reply

[Updated on: Thu, 09 May 2024 23:17]

Report message to a moderator

Re: De jure region of residence (v139) - Kenya 2014 [message #29205 is a reply to message #29202]

Fri, 10 May 2024 07:35

Bridgette-DHS
Messages: 3199
Registered: February 2013

Senior Member

Following is a response from Senior DHS staff member, Tom Pullum:

I would agree with using v024, as you proposed. In the IR file for this survey, I entered "tab v024 v139." There's no difference between the two variables, except for the last category of v139, which includes only 2.5% of women.

Report message to a moderator

Re: De jure region of residence (v139) - Kenya 2014 [message #29206 is a reply to message #29205]

Fri, 10 May 2024 09:13

dperezme
Messages: 2
Registered: October 2022

Member

Dear Tom,

Thanks for your quick response. I agree with you, comparing both variables in the KR it seems the same variable. In fact, comparing the categories of v139 and v024, it looks like they coincide in the observations except (obviously) on the not de jure resident, so the problem is that the labels are not set correctly in v139. For example, you see that the category "Kilifi" (v139=9) corresponds to Nairobi in v024, that "Kwale" would be Nyanza, and so on... By looking at in this way, you can see at the end which category of v139 corresponds to each of the Kenyan regions. Or I think it is, at least.

Thanks again for your advice.

Report message to a moderator

Previous Topic:	Birth's Recode Dataset
Next Topic:	ANC components - discrepancies with final report usingGitCode

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Nov 22 13:51:30 Coordinated Universal Time 2024