I want to find the value of parameter $k$ from my data set. The data set is composed of several populations. Should I calculate the parameter $k$ for each subpopulation, or for the population at large?
I am trying to use the fitdist(data = x, distr = "nbinom", method = "mle")
function, but I am not sure which term in this output would give the parameter $k$'s estimated value.