To all the contributors of the MDS discussion,

thanks a lot for your help. I learned a lot from that debate and got
very helpful hints, even if some aspects you discussed are beyonde my
scope. It seems there is no simple answere to what I (in the first
place) thought to be a simple question. Anyway I found the argument
concerning metric vs. nonmetric MDS helpful and of corse the various
pointers on which programs to use, their fit properties etc. Even if I
cant´t consider all your suggestions I am a great deal further to solve
that task. It might be that setting up my own code might be more than I
can do, but using available software I can still try to keep the black
box as small as possible, and not the least due to this list.

yours, dirk