by jlaake » Fri Aug 30, 2019 12:39 pm
Doesn't look like anyone answered your question. The easiest thing to do would be to use site as a grouping factor as well and then you would automatically get real estimates for each site with its se. But given that you used it as an individual covariate, then you can specify the value of the covariate you want to have MARK (assuming that is what you are using) use when it computes real parameters. You used to have to run a model with each covariate value but I think Gary changed this so you can specify a set of values to compute the real parameters without re-running. Look at the documentation and I'm sure you'll find it. I use RMark so not sure where it is.