Many engineers lack the necessary expertise to correctly analyse and interpret data resulting from an experiment. The conceptual framework of stochastics lies in the possibility of measuring the expectation of future' events on the basis of ‘past’ statistical data. Their observation leads in general to a decrease of uncertainty with respect to the initial situation: the measurement of this uncertainty can be quantitatively carried out through the concept of probability. We maintain that the precise and rigorous formulation of this process of ‘learning from experience’ can be realized by a merging of the Bayesian inferential procedure with the subjective and intuitive ‘real-life’ concept of probability as degree of belief.