[statnet_help] Modeling interaction effects in ERGM

[Pavel Krivitsky via statnet_help]

Hi, Wallace,

For mutuality, it may in fact, be well-defined, but not in general. A better way to model this type of interaction is probably via the F() (for filtering) operator.

To fit something like what you have here, you could try:


net ~ edges +

      For(l = sort(unique(net %v% "Gender"))[-1],

          F(~mutual, ~nodefactor("Gender", levels = I(l)))

      )


(I think there's an example in https://doi.org/10.18637/jss.v105.i06 .)

Now that I think about it, there's no reason F() shouldn't be able to test if the filtering term has multiple statistics and act accordingly. Maybe in the next release...


I hope this helps,
Pavel

On Fri, 2025-05-30 at 18:35 -0700, Wallace Chipidza via statnet_help wrote:
Hi,

I have a network with various demographic attributes and am testing some hypotheses on interactions between node attributes and structural effects. Is there a way to model the interaction between an attribute like gender and reciprocity in ERGM? Some way to test if women are more reciprocal in their network behavior than men? So far I've tried:

model1 <- ergm(net ~ edges + mutual:nodefactor("Gender"))

and I receive the message "Error in `ergm_Init_abort()`:! In unknown function: Change statistic interactions are poorly defined for dyad-dependent terms. Use ‘interact.dependent’ term option to set the behavior.FALSE"

Wallace
_______________________________________________
statnet_help mailing list
statnet_help at u.washington.edu<mailto:statnet_help at u.washington.edu>
http://mailman23.u.washington.edu/mailman/listinfo/statnet_help



Confidential communication - This email and any files transmitted with it are confidential and are intended solely for the addressee. If you are not the intended recipient, please be advised that you have received this email in error and that any use, dissemination, forwarding, printing, or copying of this email and any file attachments is strictly prohibited. If you have received this email in error, please notify me immediately by return email and destroy this email.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman23.u.washington.edu/pipermail/statnet_help/attachments/20250602/2feccaef/attachment.html>