Propagation of distributions by a Monte Carlo method, with an application to ratio models

The GUM uncertainty framework, namely the law of propagation of uncertainty and the characterization of the measurand Y by a Gaussian distribution (or a scaled and shifted t-distribution), is seen as an approximate implementation of a fundamental concept, the propagation of distributions. This concept and a Monte Carlo method that implements it in a numerically controlled way are outlined. A family of models, relating to comparison measurement, and solvable analytically in algebraic form, is treated by both approaches to assess the degrees of approximation involved.