I'm working with a network population of company directors. Amongst other variables, "Male"/"Female" is a categorical variable in my dataset.
I want to compute the Eigenvector centrality of the females in the network population and compare it to that of the males, however I do not want the centrality to be calculated based on filter by gender. Essentially, I want to know the influence of female nodes in the network compared to the influence of male nodes in the network, but not necessarily the influence that female nodes have within just the population of female nodes, and males within the male population. Does that make sense?
It seems that when I partition based on gender (either "Male" or "Female"), and then run statistics, the output is concerning only the partitioned population's interactions within itself. Is that right? Instead of having the options of calculating a) centrality between females, b) centrality between only males, and c) centrality within the network as a whole, how to I made a gender comparison based on the full population?
Let me know if I can clarify anything--thanks!