Hi,
I would like to ask what is the proper correlation coefficient to use for count data and for zero inflated count data.

would Kendall tau do the job or would the data violate some assumptions of it?