Martina Morris

Sure -- if there are no copyright issues, we'd be honored to do that.

We're in the process of updating the statnet website -- into a GitHub hosted version. Would be happy to post it there in a "Resources" section if you like. At...

@smjenness CRAN. The issue, as you know, is reducing the maintenance when the calculation of terms is modified. There may be no downside to keeping a couple of terms inside...

We should also discuss how this might link to the `ergm.ego` workflow.

This is a good question @sgoodreau. Everytime I've ventured into this from a research perspective, I've wanted to be able to decompose the overall variation into within and between simulations....

Right, as soon as you mentioned the imputation for censored start times I remembered. I'll think on this.

I agree that the SE is useful for capturing the variation in the estimate of the mean -- though its statistical interpretation depends on whether one knows the sampling distribution...

@chad-klumb that's a useful reminder, thx. I thought the formula was p $\approx$ 1/(SE^2) rather than 1/(z^2). The approximate p-value that Chebyshev's inequality gives can be quite conservative (larger) relative...

This is looking really good. Lots of info, should cover most user contexts now. One q about naming for the SD: it's not clear to me what the "1-" in...

lol, ok, let me think on it a bit