Re: factor analysis /print det problem? on Mac

Posted by David Marso on Jun 19, 2013; 12:32am
URL: http://spssx-discussion.165.s1.nabble.com/factor-analysis-print-det-problem-on-Mac-tp5720804p5720805.html

ODD!
Looking at the algorithms it would seem that KMO would be undefined (don't we need R inverse for AntiImage matrix) and Bartlett 0.000 in the presence of a 0 determinant.  Mind doing a MATRIX out on the factor and attaching the correlation matrix/means/SDs (MATRIX SYSTEM FILE).
--

Maguin, Eugene wrote

For people that are using v20 (release 20.0.0) on Macs, have you noticed/has it been noted (not it matters much now that 22 is being sold) that the print det (determinant) yields a blank entry in the determinant object? The background is that somebody did a factor analysis and requested the det but there is not det value but the label is superscripted that the determinant is 0. However, I don't really believe that because the KMO is .740 and Bartlett is p=.000 and rotated pattern matrix has loadings in the .7 range. So, I'm curious.
Thanks, Gene Maguin

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD