The DHS Program User Forum: Sampling

Home » Data » Sampling » Sampling (Relative Margin of Error (RME))

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

Sampling [message #29082]

Mon, 22 April 2024 09:00

Nahid
Messages: 9
Registered: November 2020

Member

Dear Concern,
What is the acceptable lower and upper limits for MRE while calculating sample size for subnational level? Many thanks in advance for your kind response.

Kind regards
Nahid

[Updated on: Mon, 22 April 2024 09:01]

Report message to a moderator

Re: Sampling [message #29093 is a reply to message #29082]

Tue, 23 April 2024 07:55

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

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

The level of the Mean Relative Error varies from one indicator to another. DHS surveys are generally designed to have acceptable levels of MRE for strata, which are usually combinations of urban/rural residence crossed with the first admin level, or regions. Just a few surveys are designed for acceptable estimates at level 2, or districts. We recommend that you calculate confidence intervals for estimates, including adjustments for the survey design, and make your own judgement about acceptability, rather than using some arbitrary cutoff such as 20%.

Report message to a moderator

Previous Topic:

Two or three stages?

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Tue Apr 1 08:12:39 Coordinated Universal Time 2025