home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!wupost!darwin.sura.net!paladin.american.edu!auvm!UMDNJ.EDU!JASKIR
- Mailer: Elm [revision: 66.36.1.1]
- Message-ID: <SPSSX-L%92072811074936@UGA.CC.UGA.EDU>
- Newsgroups: bit.listserv.spssx-l
- Date: Tue, 28 Jul 1992 11:07:41 EDT
- Reply-To: John Jaskir <jaskir@UMDNJ.EDU>
- Sender: "SPSSX(r) Discussion" <SPSSX-L@UGA.BITNET>
- From: John Jaskir <jaskir@UMDNJ.EDU>
- Subject: Final Statistics in Factor
- Comments: To: spssx-l@uga.cc.uga.edu
- Lines: 18
-
- Can anyone shed some light on why SPSS-PC does not compute final
- eigenvalues and percent of variance explained for rotated factors ?
-
- Specifically, I'm refering to a principal components analysis with
- varimax rotation.
-
- The "final" statistics in the output, I believe, are potentially
- misleading since it refers to the variance explained by the factors
- only after the extraction stage. However, when factors are orthogonally
- rotated the distribution of the explained variance changes amoung the
- factors. The only way I've been able to determine this new distribution
- is to calculate the eigenvales by hand and divide by the number of
- variables.
-
- Am I missing something?
-
- ----------------------------------
- jaskir@umdnj.edu
-