I am working with a database that involves weighted statistics and complex survey design, namely the National/Nationwide Inpatient Sample, so I am using R's 'survey' package for tasks like summary statistics, regression, etc.
One test I was interesting in running was comparing the population of a certain group in 2 different populations (i.e. is the difference between the proportion of A in Population B and proportion of A in Population C statistically significant). I have already used scvyciprop to plot the confidence intervals for these proportions, and have seen that these two proportions are significantly different. However, I was wondering if there was a function like prop.test, whether or not it's in the 'survey' package, that can run this test for data in a complex survey (ex. takes a survey design object as an argument) and actually quantify the p-value / t-statistic. Does svyttest have this functionality, or is there another function I could potentially use?
Thank you!
