New approach for multivariate analysis of EM data MsDEMPCA (Missing data EM Principal Components Analysis) is applied to several large arrays: the Baltic Electromagnetic Array Research (BEAR), Electromagnetic Mini Array (EMMA), Magnetotellurics in Scandes (MASCA), IMAGE obesrvatory data and various EarthScope MT arrays, with the goal of demonstrating how the MV approach can clarify signal and noise characteristics, improve estimates of TFs, and, perhaps most importantly, quantify biases in these estimates.