Dear list, Apologies if this turns out to be a silly question but does predict.MCMCglmm, marginalise residuals by default? I'd like to see how the prediction performs when random effects are not marginalised, i.e., marginal=NULL, but the predictions look too close to the original data so I am suspecting that this also un-marginalised the residuals as well... Am I correct to interpret it this way, and is there a way to still marginalise the residuals but not the random effects? I want to see how close or how far away the predictions are from the original responses when random effects are marginalised or not. Many thanks, Manabu
Manabu Sakamoto, PhD manabu.sakamoto at gmail.com [[alternative HTML version deleted]]