Generalized Linear Mixed Models (Vid 4)

A bit on standard error, Monte Carlo standard error, variance components, and a quick comparison to other methods. MCEM finds the MLE. We want results to match the point estimates produced by MCEM. We can't rely on just MCEM though because it only produces MCMLEs; MCLA approximates the entire likelihood so we can do any frequentist likelihood-based inference (likelihood ratio tests, Wald tests, confidence intervals, etc). Thanks for tuning in! This will be the last video in this series...maybe. I may come back to fill in and flesh out a few more details. I kind of want to do a video on Wald tests. We'll see if I have time!