This paper emphasizes that numerically correct calculation of economic uncertainty with intervals and fuzzy numbers requires implementation of global optimization techniques in contrast to straightforward application of interval arithmetic. This is demonstrated by both a simple case from managerial economics as well as a real life railway reconstruction project. Based on identical uncertain input data the difference between the probabilistic and the possibilitistic approach is highlighted, the latter producing a substantial larger degree of numerical uncertainty due to its non-statistical nature. 2010 World Academic Press, UK. All rights reserved.
Journal of Uncertain Systems, 2010, Vol 4, Issue 1, p. 47-58