I have a data set of 30 items, each with its own length, height, and weight. I'd like to create 3 groups where the average length, height, and weight are as close to identical as possible across those 3 groups. You can think of this as creating k-clusters, but rather than maximizing the differences between the groups, I want to eliminate the differences. Is there a way to do this in SPSS?

------------------------------
Danielle Gunraj
------------------------------

#SPSSStatistics

Hi. I am no expert in this area, but have you looked into the PROXIMITIES procedure, with its many similarity and dissimilarity measures? Perhaps you could use one of those, save out the matrix of similarities (or dissimilarities, as the case may be), and input that matrix to CLUSTER. Again, I have not done that, but it's the first place I would look if I needed to.

------------------------------
Rick Marcantonio
Quality Assurance
IBM
------------------------------