Refinements to the variety metric for idea evaluation

Systematically developing innovative products demands effective idea generation methods. The effectiveness can be verified based on a set of metrics, of which the variety metric is one. It is demonstrated that this metric exhibits several shortcomings, such as the lack of level-based measurements, and arbitrarily defined level weights. The currently applied variety metrics, furthermore, do not measure the degree of uniformness of the distribution of ideas over nodes on an abstraction level. A level-based, correctly normalized variety metric which accounts for the degree of uniformness of the distribution of concepts over nodes, is proposed, and is shown to resolve the above issues. (C) 2012 Elsevier Ltd. All rights reserved.