[Rsiena-help] interpretation of the outInv parameter in 2-mode networks

[marco tonellato]

Dear All,

while trying to model a 2-mode network with a very skewed degree
distribution of the actors of the first mode, I ran a model with the
"1/outdegree+c" effect (outInv) and got a very large positive but very
significant coefficient (theta=18 stdev=1.8 ), with convergent t-ratios. The
outdegree activity effect (outActSqrt) instead gives convergent results but
extremely high standard deviation.
I would like some help in the interpretation of the outInv parameter in
general and of the large coefficient in particular. Can I still interpret
outInv as a parameter capturing the degree distribution, similarly to
outActSqrt?

Thanks in advance.
Best,

Marco
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.r-forge.r-project.org/pipermail/rsiena-help/attachments/20100330/793f886c/attachment.htm