One of the many recent developments in statistical modeling of social networks has been the so-called p2 model. I will present this random effects model and place it in context with other available models for complete network data.
With MCMC methods good estimates of the parameters of the p2 model can be obtained. It seems more difficult to find good ways to select models and to assess their goodness-of-fit, two topics I would like to bring up for discussion.
