Date: Thu, 2 May 1996 09:15:54 -0400
Reply-To: "Peter H. Hartmann, FI" <phhart@WISO-R610.WISO.UNI-KOELN.DE>
Sender: "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From: "Peter H. Hartmann, FI" <phhart@WISO-R610.WISO.UNI-KOELN.DE>
Organization: Uni Koeln, Seminar fuer Soziologie
Subject: Pairwise deletion with cluster / proximities
Indeed it is a problem that it is not possible to use pairwise deletion
of missing values with proximities. I think it would be easy to include
such an option into the proximity module and I cannot understand why this
thing only does listwise and include! Consider, e.g., that correlation
uses pairwise as default, so why should it be impossible in proximities.
When, e.g., you use proximities to generate matrices of Yule's Q coeffi-
cient (I see no other way of doing this than using pairwise), it usually
makes a lot of sense to use pairwise deletion!
Peter H. Hartmann
Univ. of Cologne