The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Kenya » 2014 Kenya Male Survey Representative at the County Level?
Re: 2014 Kenya Male Survey Representative at the County Level? [message #9805 is a reply to message #9794] Mon, 23 May 2016 08:37 Go to previous messageGo to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3230
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum:

Unless specified otherwise in the main report, all DHS samples are representative at all levels, in the sense that the estimates (weighted appropriately) are unbiased. Some old documentation on this says the contrary, but in fact that statement applies to all DHS surveys. As you go down to smaller areas, the estimates remain unbiased, but the standard errors can become quite large because the number of clusters gets smaller and the clustering effect becomes more serious. Confidence intervals get wider. The intervals will be larger in some of the counties than in others.

I would just recommend that you pay attention to the confidence intervals of the county-level estimates, especially for the counties with fewer clusters. You may want to set a cutoff such that you refuse to provide an estimate if the confidence interval is so wide that the estimate is virtually useless. This would be analogous to how DHS just gives an asterisk if a proportion is based on fewer than 25 cases, and puts an estimate in parentheses if it is based on 25 to 49 cases.
 
Read Message
Read Message
Read Message
Previous Topic: why weights cannot be used to test for relationship like regression
Next Topic: 2014 DHS Kenya Male BMI variable
Goto Forum:
  


Current Time: Tue Feb 25 09:01:26 Coordinated Universal Time 2025