Print

Print


Dear SOCNET-ers,

I have a question about Bonacich Power and the Centrality/Power routine in
UCInet. My research is on the role of state ownership in three countries.
Hence, I have three separate networks, one for each country. Lets say that I
want to make comparisons about the relative role of state ownership in the
three countries. Is it valid to compare the actual power value for the State
Treasury of Country X in the network of Country X, against the power value
of the State Treasury of Country Y in the network of Country Y? For example,
if the value for the state treasury in country X is 150 and the value for
the state treasury in country y is 50, can I assume that state ownership is
approx. 3 times as powerful in the network of country X than the state in
the network of country Y?

If so, how strongly does the value of Beta (different values were used in
the three routines in accordance with the "Beta should be no larger than
1/the largest eigenvalue of the matrix" rule suggested by Bonacich) affect
the actual values of power in the output? Does this pose problems for
comparison across networks?

Many thanks in advance.

Roger Schoenman