The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » General » DHS STATcompiler vs DHS household data (difference in marital status data)
DHS STATcompiler vs DHS household data [message #20504] Thu, 12 November 2020 12:10 Go to next message
nbotev is currently offline  nbotev
Messages: 2
Registered: November 2020
Member
I am working with UN Population Division's World Marriage Data 2019 (https://population.un.org/MarriageData/Index.html#/home). Part of the marital status data come from DHS. In this case, the dataset has separate entries for "DHS STATcompiler" and "DHS household data". The differences between the two sources are small, but still the UN Population Division staff have opted to include both in the dataset. Where do the differences come from, and which of the two sources would you recommend to use?
Thank you in advance for your help with that issue.

[Updated on: Thu, 12 November 2020 12:11]

Report message to a moderator

Re: DHS STATcompiler vs DHS household data [message #20505 is a reply to message #20504] Thu, 12 November 2020 16:43 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2345
Registered: February 2013
Senior Member

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

The household survey includes a roster in which every member of the household is listed and then some questions are asked about each person. One of these is marital status. The responses are given by the household respondent, who is usually with the household head or the spouse of the household head, but may be someone else so long as they are age 15 or above. For example, it could be a grandparent or a grown child.

Then the women and men who are eligible and are interviewed are asked directly about their marital status. That's the main source of the difference--whether the information comes from the household respondent, or directly from the adult woman or man. You will see a similar difference in the distribution of educational level. There are sometimes differences in reported age.

There are occasionally differences even when the household respondent and the individual woman are the same person, perhaps because there is somewhat greater clarification of categories by the interviewer during the individual interview. Also sometimes I think there is genuine ambiguity in what is the "true" marital status, especially in a context where "marriage" may not involve a formal process.

Re: DHS STATcompiler vs DHS household data [message #20513 is a reply to message #20505] Fri, 13 November 2020 07:34 Go to previous message
nbotev is currently offline  nbotev
Messages: 2
Registered: November 2020
Member
Thank you so much, this is very helpful.
Previous Topic: Merge Service Availability data with individual data for zimbabwe
Next Topic: information on average number of households per enumeration area in Mozambique 2011 DHS
Goto Forum:
  


Current Time: Sat Jul 24 03:37:46 Coordinated Universal Time 2021