Model validation and vec operators
Cristina Dias, Carla Santos, Maria José Varadinov, João Tiago Mexia · AIP conference proceedings · 2018
We use the vec and other relater operators to carry out inference for structured families of symmetric stochastic matrices M. These are obtained through the sum of the respective mean matrix and a symmetric stochastic matrix with null mean. We consider that the vec operator of the matrix E is normal homoscedastic. The matrices on these families correspond to the treatments of a base design, and the inference is centered on model validation and the action of the factors in the base model on mean matrices.