The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » General » Replacements? (Survey Protocols for Unavailable Households)
Replacements? [message #19891] Tue, 25 August 2020 10:42 Go to next message
sylvan is currently offline  sylvan
Messages: 4
Registered: August 2020
Member
I haven't been able to find any documentation or associated variables for replacement households. I know DHS works hard to get the targeted/initially listed households, but were there ever replacements if a household was untrackable, unavailable, or unwilling to participate? 100% of households are listed as "completed" which can't be right.

On an individual level, what happened if an eligible male/female were not available? Were they excluded? Is there a record of this?

Thanks!
Re: Replacements? [message #19894 is a reply to message #19891] Tue, 25 August 2020 10:49 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1965
Registered: February 2013
Senior Member
We don't distribute households or Individuals with result of interview different from 1 (code 1 represents a completed interview). The reason is because that information is only used to calculate response rates, which are published in all DHS final reports. We don't collect any information for households/individuals with result of interview different from 1 (completed).
Re: Replacements? [message #19895 is a reply to message #19894] Tue, 25 August 2020 10:54 Go to previous messageGo to next message
sylvan is currently offline  sylvan
Messages: 4
Registered: August 2020
Member
Thank you Bridgette. That's what I inferred from the data... Two quick follow-ups:

1) Is there any way of knowing which households were replacements?
2) Is there any way of knowing response rates by enumeration area?

Thank you!
Re: Replacements? [message #19897 is a reply to message #19895] Tue, 25 August 2020 14:10 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1965
Registered: February 2013
Senior Member

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

There is nothing in the data files that tells whether a household is a replacement. This is tracked during fieldwork but nothing goes into the data files. As you say, replacements are rare. More often, the household is just omitted and not replaced and the household weight is adjusted to compensate (at the stratum level)

However, you can easily check individual-level nonresponse. Merge the IR file with the eligible women in the PR file, i.e. the PR cases with hv117=1. Cases with _merge=1 or 2 (depending on the sequence of the files in the merge, only one of these codes will appear) are nonresponse cases. Similarly for men, using the MR file and PR cases with hv118=1. You can then calculate nonresponse rates by clusters, interviewers, etc.
Re: Replacements? [message #19898 is a reply to message #19897] Tue, 25 August 2020 14:23 Go to previous message
sylvan is currently offline  sylvan
Messages: 4
Registered: August 2020
Member
Thank you so much Bridgette. This is super helpful!

Very best, Sylvan
Previous Topic: Strata, PSU, weights for Sengal Continuous Survey
Next Topic: svyset-Multilevel modeling
Goto Forum:
  


Current Time: Sat Oct 24 10:45:15 Eastern Daylight Time 2020