5
$\begingroup$

After I multiply imputed my dataset m times I wanted to calculate a binomial proportion confidence interval. I did that formerly using the Hmisc::binconf() function in R, but pooling seemed to be impossible using the mice package. How do I combine these confidence intervals into one considering within and in-between imputed dataset variance? Can I use Rubin's rules?

Update:

After reading the article mentioned in the comments, I may have found a way to pool binomial proportion confidence intervals. First I used the Wilson Score Interval formula:

Wilson

Then I calculated the relocated centre estimate p' using the following formula for every complete dataset:

Center Wilson Interval

And after that I calculated the correct standard deviation s' for every complete dataset using a formula found in this article:

Standard deviation

By calculating the standard deviation and relocated centre for every complete dataset I was able to pool the results using mice::pool.scalar(Q, U, n = 100), and calculate the final confidence interval using the qnorm(0.975, Q, sqrt(U) function. Can anybody confirm whether this is correct or not?

$\endgroup$
  • $\begingroup$ For a general idea of Rubin's combining rules, you can read ncbi.nlm.nih.gov/pmc/articles/PMC2727536. The original book is eu.wiley.com/WileyCDA/WileyTitle/productCd-0471655740.html $\endgroup$ – Qaswed Jun 22 '16 at 18:29
  • $\begingroup$ @Qaswed I rephrased the question as it might have been too broad. Thanks for the article, can you have a look at my updated question, I hope I managed to do it correctly now. $\endgroup$ – Roger Jun 23 '16 at 13:50
  • $\begingroup$ @Roger Although the score interval formula is a good way of calculating confidence intervals around a proportion, you could also try using the easier to apply Agresti-Coull adjusted Wald confidence intervals (see link ) The problem however, lies in taking into account the variance between imputation datasets. Neither the Score or Agresti-Coull CIs take this into considerations. I have upvoted your question as I'd also like an answer to this, but as to whether or not using only Score CIs is correct, I'd say no. $\endgroup$ – IWS Jan 11 '17 at 13:16
2
$\begingroup$

This is indeed an interesting problem. The issue is that the standard errors that are based on the central limit theorem for proportions are often undesirable because proportions are a computed quantity and for that reason exhibit skewed uncertainty over sampling. The Wilson score, such as you mentioned, gets around the skewness by estimating a different quantity than the standard proportion $k/n$. What you need to use Rubin's rules is an estimate of the within-imputation variance of this transformed proportion, which is just the variance/standard error estimated on a single dataset, along with the transformed proportion itself for each dataset.

So for the Wilson score interval, you first need to calculate the transformed estimate $ \hat{p} + \frac{1}{2n}z^2 $ and then separately the variance, which from your formula is $ (z\sqrt{\frac{1}{n} \hat{p}(1-\hat{p}) + \frac{1}{4n^2}z^2})^2 $

That will give you estimates of the transformed parameter and the transformed parameter's variance for each of $m$ datasets.

You can then combine these estimates using some of the available R tools, such as mi.meld from Amelia or mice as you mentioned or the R package mitools. Then once you have the transformed parameters, you can compute the confidence interval based on the newly derived variance/parameter estimate.

This would be easier if these R packages supplied the transformed estimates instead of just the confidence intervals, but you can probably dig them out of the associated R code.

$\endgroup$

Your Answer

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

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