I would like to use R to calculate the standard error of the median income of a population represented in PUMS. I would like to do this using replicate weights, if possible.
I know how to calculate the standard error of an estimate using replicate weights, but the calculation of median income seems trickier. How do you properly weigh that type of data? Each person's income is weighted differently so... ?
I would like to stick with R and replicate weights if at all possible (ie no SAS, SPSS, or design factors)