On identification in conditional models

A general definition of a conditional model is proposed through embedding that concept into the usual concept of a standard statistical model. The identification of a conditional model is eventually analysed in the framework of the identification of a function of the parameters in a standard statistical model. A simple example shows some practical implications of the proposed approach. Whereas this paper is written in a pure sampling theory framework, a final remark connects this framework with the bayesian approach to the identification in conditional models