Closed wcornwell closed 5 years ago
I just re-invented DFBETA here:
https://cran.r-project.org/web/packages/olsrr/vignettes/influence_measures.html
Agreed. Saw that page and thought DFBETA might be useful.
What we could potentially do would be to build a model with 'space and time' and then get DFBETA and see how that parameterizes the formula.
Cook's D is for the entire model, so might be worth considering a modification (or addition) to look at the influence of each point on the SE of the specific terms that are crucial for IUCN red list criteria.