1
$\begingroup$

I work in epidemiology on a sample which is stratified, and 2 – level cluster (801 individuals). I use the "survey" package on R for data analysis.

In the sample, the sum of weights is the size of estimated population.

When I calculate prevalence in sample 1 with weights or normalised weights, the results were the same, but I thought the confidence interval could be impacted and the variance too. Why it is not the case?

Thanks you in advance,

$\endgroup$

1 Answer 1

1
$\begingroup$

The use of normalized weights is done to reduce extreme weights on one end or the other of the spectrum. It's possible that, for your data, no or few such weights occurred, so the normalized weights and non-normalized weights give very close to the same answer.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.