Estimating Uncertainty of Correlated Quantities - Comparison of Two Methods

The paper considers uncertainty estimating when uncertainty sources are correlated and probability density function distribution is not normal. Two approximate methods for evaluating uncertainty are compared. The first is based on recommendations in "Guide to the Expression of Uncertainty in Measurement". The second one uses Reductive Interval Arithmetic [2]. As a reference there is used an uncertainty calculated basing on its definition and simulation generated random variable.