I2 uncertainty estiamtion #10
Description
Currently, use Monte Carlos simulations to get SE (probably should use n -1 rather than n; n being the number of levels). It would be good to compare ones from MCMCglmm and also, compare with ones obtained from parametric bootstrapping. I shall look into this soon while I work on something else (e.g. GrptR).