A new bibliometric index is proposed, trying to preserve the advantages of the h-index and to overcome its disadvantages. Multivariate comparisons among 18 bibliometric indices are performed by using Hasse Diagram Technique (HDT) and Principal Component Analysis (PCA). The comparisons were performed on some artificial data sets, three of them well known in literature. The obtained results seems to highlight some interesting properties of the new index and also reveals some relevant relationships among the considered bibliometric indices.