Question

I have 2 questions for survey data:

1, What's the difference between svy:tab and svy,subpop(XX): tab? if subgroup first then use svy:tab.

2, Do we have the corresponding SAS code as svy, subpop(XX): tab?

Thanks in advance for your help and time.

Was it helpful?

Solution

The subpop() option means that estimates are only computed for a single subpopulation identified by XX not equal to 0 and not missing. For a simple binary XX variable, that would be equivalent to XX==1. The reason to use subpop() is that merely adding an "if XX==1" restriction will often fail at producing the correct variance estimates for the subpopulation. However, if you're just looking at proportions with tab, that should not make a difference since these are point estimates. You can find more details (including formulas) here. I am not sure what the equivalent SAS code is.

Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top