LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (February 1998)Back to main SPSSX-L pageJoin or leave SPSSX-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Wed, 25 Feb 1998 08:00:00 PST
Reply-To:     user spss <user_spss@BANDMASTER.PORTAL.CA>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         user spss <user_spss@BANDMASTER.PORTAL.CA>
Organization: Another WaterGate site !
Subject:      Re: HELP: Cluster Analysis in SPSS

Mauri Ziff wrote: > > e-mail me at: mauri_ziff@quickmail.ucsf.edu > > Subject: Time:4:50 PM > OFFICE MEMO cluster analysis > Date:2/19/98 > > I am interested in doing a cluster analyses (CA) using SPSS for a > study on older adults and behavior. It is unclear to me whether it is > possible to do a CA if the data include different types [i.e., the > behavior variables are binary (dichotomous), continuous, and ordinal, > and I want to use them in the same analysis]. And, if it is possible, > which distance/proximity measure would be the most appropriate to use? > > Thank you. > M. Ziff

SPSS's two cluster procedures, CLUSTER and QUICK CLUSTER, require all variables to be treated as interval-level. However, true dichotomies (coded as 0 or 1 for the presence or absence of an attribute) can be regarded as interval, and most ordinal scales can also be treated in the same manner with little risk of significant error. Of course, it all depends on the nature of your variables.

CLUSTER, useful for relatively small datasets, admits a number of distance measures. QUICK CLUSTER uses only Euclidean distance. In both cases, variables are standardized with zero mean and unit variance.

Hector Maletta Universidad del Salvador Buenos Aires, Argentina

Hector Male

-- +----------------------------------------------------------------------+ | This message was gated by The BandMaster BBS -- FidoNet 1:153/7715 | | with the assistance and cooperation of Internet Portal, Vancouver BC | | Ain't FidoNet Great? :-))) | +----------------------------------------------------------------------+

-- +----------------------------------------------------------------------+ | This message was gated by The BandMaster BBS -- FidoNet 1:153/7715 | | with the assistance and cooperation of Internet Portal, Vancouver BC | | Ain't FidoNet Great? :-))) | +----------------------------------------------------------------------+

-- +----------------------------------------------------------------------+ | This message was gated by The BandMaster BBS -- FidoNet 1:153/7715 | | with the assistance and cooperation of Internet Portal, Vancouver BC | | Ain't FidoNet Great? :-))) | +----------------------------------------------------------------------+

-- +----------------------------------------------------------------------+ | This message was gated by The BandMaster BBS -- FidoNet 1:153/7715 | | with the assistance and cooperation of Internet Portal, Vancouver BC | | Ain't FidoNet Great? :-))) | +----------------------------------------------------------------------+

-- +----------------------------------------------------------------------+ | This message was gated by The BandMaster BBS -- FidoNet 1:153/7715 | | with the assistance and cooperation of Internet Portal, Vancouver BC | | Ain't FidoNet Great? :-))) | +----------------------------------------------------------------------+

-- +----------------------------------------------------------------------+ | This message was gated by The BandMaster BBS -- FidoNet 1:153/7715 | | with the assistance and cooperation of Internet Portal, Vancouver BC | | Ain't FidoNet Great? :-))) | +----------------------------------------------------------------------+


Back to: Top of message | Previous page | Main SPSSX-L page