Date:         Tue, 22 Apr 1997 10:32:45 -0500
Reply-To:     chelberg@spss.com
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         Clay Helberg <chelberg@SPSS.COM>
Organization: SPSS, Inc.
Subject:      Re: Error Messages in LISREL 8
Content-Type: text/plain; charset=us-ascii

Michelle Yik wrote:
>
> I got the following 2 messages in a row.  Is there any way to get around
> with that?  Any comments are welcome and please direct all correspondence
> to myik@unixg.ubc.ca
>
> Thanks in anticipation!
>
> W_A_R_N_I_N_G: Matrix to be analyzed is not positive definite,
>                  ridge option taken with ridge constant =  10.000
>
> F_A_T_A_L  E_R_R_O_R: Matrix to be analyzed is not positive definite.
 
Are you using pairwise treatment of missing values? This can often lead
to NPD correlation/covariance matrices. If so, you might have better
luck using listwise deletion of missing values. If you have too much
missing data, you will have to deal with it, either by collecting more
data, dropping variables, or using a mathematical solution such as
imputation of missing values.
 
                                                --Clay
--
Clay Helberg        | Internet: chelberg@spss.com
SPSS, Inc.          | WWW: http://www.execpc.com/~helberg
444 N. Michigan Ave.|
Chicago, IL 60611   | Speaking only for myself....