More and more frequently, different materials have to be compared to find out the most suitable for a specific use. A large number of individual properties and performances have to be considered at the same time and a synthetic evaluation via a global performance index (GPI) may be very useful. In this paper, a new methodology, previously used for durable goods, is applied for the first time to polymeric materials. The GPI can be obtained, starting from basic parameters expressing the performance of each material considered, through the construction of a 'relevance tree' with increasing level of data aggregation. The construction can be performed either graphically or algebraically by using nomograms or a computer implemented algorithm, respectively. In the latter case, the influence of the averaging functions over the final GPI is also discussed.
