Comments on testing economic theories and the use of model selection criteria