In many cases, the quantitative relevance of physical effects for a given technical problem is not known a priori. This holds especially for the analysis of the dynamics. Adopted from nonanalog circuit design, in the last years symbolic model reduction techniques found their way towards mechatronic system modeling. Given a scenario (system inputs, initial values, parameters) and an error bound, symbolic model reduction reduces the detailed model to a less complex model, which is guaranteed to stay within predefined error bounds. However, presently symbolic reduction techniques deliver reduced models, which are only verified for a single scenario. For example a reduced vehicle model emerging from the reduction of a complex multibody vehicle model for a cornering maneuver with a small constant steering angle, is not verified to stay inside the error bounds for any other maneuver. In this contribution this drawback is addressed by the use of interval-valued scenarios.