I am familiar with Rubins rules for combining results from multiple impuations to get a regression estimate and variance.

I was wondering, when estimating a proportion bound from 0 to 1 from multiply imputed data, does Rubins rules still apply?

Also, if I wanted to calculate Clopper Pearson, Korn Graubbard or another type of non-symetric confidence interval that doesn't use the standard error in the equation, how would I do this? Would I calculate the lower point and upper bound the same way I would to calculate the point estimate?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.