SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to
Dear Ian and others,
To use this option in RSiena where the rate parameter of the actor variable is set to something approximating minus infinity (-100 will do), it will be helpful to use a non-centered dummy variable (1 for M, 0 for F).
For this kind of variable (pregnancy), where a duration / event history model would be appropriate, the model developed in
Greenan, Charlotte C., (2015).
Diffusion of Innovations in Dynamic Networks.
Journal of the Royal Statistical Society, Series A
, 178, 147-166.
might be a good idea.
See also section 12.2.4 in the RSiena manualhttp://www.stats.ox.ac.uk/~snijders/siena/RSiena_Manual.pdf
This assumes variables that only can go up (diffustion of innovation), but it could also be applied to other variables for which duration modeling is meaningful. Then, however, also the current state (pregnant / non-pregnant) should be used as a predictor for the rate function.