*****  To join INSNA, visit  *****

HI, Emily,

The ergm.ego framework that Michal had linked to should still be applicable even if there is alter overlap, because it shouldn't bias the estimates of ERGM sufficient statistics under (hypothetical) repeated egocentric sampling, which is what ergm.ego is based on.

In addition to what Michal and Raffaele had suggested, if you have identifiable alters, you may also be able to use the missing data approach of
Handcock, M. S. & Gile, K. J. (2010) Modeling Social Networks from Sampled Data. Annals of Applied Statistics, 4(1): 5-25. doi:10.1214/08-AOAS221
or its Bayesian equivalent of
Koskinen, J. H., Robins, G. L. & Pattison, P. E. (2010) Analysing exponential random graph (p-star) models with missing data using Bayesian data augmentation. Statistical Methodology, 7(3): 366 - 384. doi:10.1016/j.stamet.2009.09.007
The former is implemented in the ergm package and the latter in PNet (I think), but if the fraction of unobserved relations is high, it may be computationally impractical.

I hope this helps,

On Wed, 2019-12-11 at 21:00 +0000, Emily Long wrote:
***** To join INSNA, visit  *****
Hi all,

We are in the beginning stages of an ego-net analysis and are interested in using ergm.ego. We have approximately 100 egonets (sadly no alter-alter ties) and due to the study design, also have a handful of alters that were nominated by more than one ego.

Is there a way to account for this in ergm.ego? Disclaimer: I'm quite comfortable with whole network analyses, but am new to ego-networks in general, so I'm not sure if the presence of common alters is actually relevant in our study. Conceptually, we are interested in the extent to which various ego attributes relate to the type of social relationships they report.

Any thoughts greatly appreciated. We are also open to other modeling options, but at the moment have a few simple ergm.ego models up and running (and so far are fun to learn).



SOCNET is a service of INSNA, the professional association for social
network researchers ( To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.