What is the best way
to test the extent of homophily across two different networks? I can use the
ANOVA density models in UCINET to test whether ties are more common among
individuals with a specific attribute. But what if I want to compare the extent
to which ties are more common among individuals with a specific attribute in
two different networks (i.e., in which network is there more homophily?). The networks have the exact same number of nodes, but
not the same ties.?Does it make sense to compare the R-squared for the models for the two networks? Or should I compare the regression coefficients from a ¡°constant
homophily¡± model for each network, for example? Would comparing either the R-squared or the regression coefficient tell me in which network there is more homophily? Thank you in advance for any advice.
