1

I am analysing data using the survey package.

There is a function svyciprop that allows for CI calculation but I am trying to find a proper way to calculate standard errors for proportions.
Only two things worked until now, using SE(svyciprop(formula, design)) or svymean(formula,design), both calculations display the same SE result, are they proper ways to calculate SE for proportions?

Marc Le Bihan
  • 2,308
  • 2
  • 23
  • 41
Jorge A. Rios-Duarte
  • 11
  • 1

1 Answers1

0

Yes, they calculate the standard error for the proportion. The standard error isn't always that useful for proportions, because you often want an asymmetric confidence interval such as the ones svyciprop computes. But if you do want the standard error, that is how to get it.

Thomas Lumley
  • 1,893
  • 5
  • 8