The greatest problem for empirical analysis is how best to allow the context to affect the inferences. Econometric theory presupposes contextual "restrictions" that can be taken as given or assigned a probability distribution. These contextual inputs are rarely available. I illustrate this point...