I'm looking for some papers on ditribution of Pearson's correlation estimator. There are very few results regarding that found with Google and same here unfortunately. The only reasonable article I found is from Wikipedia:
https://en.wikipedia.org/wiki/Fisher_transformation
However derivation of Z statistic is not exhaustive enough for me. Also, notice how the mean of distribution of Z depends on the real correlation coefficient. Thus it's of no practical use. Using the estimator value instead would be plain wrong, so we're still left virtually with nothing.
Some time ago I had to estimate correlation between two samples assumed to come from normal distribution each. Also, it was assumed they form bivariate normal with all paramerters uknown. This is the framework I expect to encounter most of the times and anything within this assumptions would be greatly aprreciated.
I guess it's probably not trivial, so please let me know of any reasonable sources where I can look up something helpful.