Re: Factor Analysis - Insolvable Varimax Rotation in SPSS
Bruce Weaver
> In an attempt to complete a text book problem on factor analysis, I
> obtained the following solution using SPSS.
>
> "Attempted to extract 2 factors. In iteration 25, the Hessian was not
> positive definite. Extraction was terminated."
>
> This illustration is problem 9.20 of Johnson and Wichern (2002),
> Applied multivariate analysis. The solution manual for the text shows
> that the problem is solvable. Would you happen to know why I did not
> obtain a solution.
>
> Your response will be appreciated.
>
> JM
>
I believe 25 iterations is the default setting in SPSS. Have you tried
increasing the iterations?
By the way, the SPSS newsgroup is a better place for questions about
SPSS software issues. I've cross-posted to that group.
--
Bruce Weaver
bwe...@lakeheadu.ca
www.angelfire.com/wv/bwhomedir
stata...@yahoo.com
Richard Ulrich
> In an attempt to complete a text book problem on factor analysis, I
> obtained the following solution using SPSS.
>
> "Attempted to extract 2 factors. In iteration 25, the Hessian was not
> positive definite. Extraction was terminated."
"not positive definite" means that it isn't a proper
covariance matrix. Most likely, there is a data entry error,
in the textbook or in your entry of it.
>
> This illustration is problem 9.20 of Johnson and Wichern (2002),
> Applied multivariate analysis. The solution manual for the text shows
> that the problem is solvable. Would you happen to know why I did not
> obtain a solution.
--
Rich Ulrich, wpi...@pitt.edu
http://www.pitt.edu/~wpilib/index.html