Correlation Matrix Input to SPSS
MATRIX DATA VARIABLES=varlist
[/FILE= {INLINE**}]
{file
}
[/FORMAT=[{LIST**}][{LOWER**}][{DIAG**}]]
{FREE } {UPPER } {NODIAG}
{FULL
}
[/SPLIT= varlist] [/FACTORS= varlist]
[/CELLS= number of cells][/N= sample size]
[/CONTENTS=[CORR**][COV][MAT][MSE][DFE][MEAN]]
[SD][PROX][STDDEV][N_SCALAR][N_VECTOR][N]
[N_MATRIX][COUNT]
**Default if the subcommand is omitted.
This worked
MATRIX DATA VARIABLES=ROWTYPE_ female race ses schtyp.
BEGIN DATA
MEAN 0 0 0 0
STDDEV 1 1 1 1
N 200 200 200 200
CORR
1
CORR -.02425329
1
CORR -.17389495
.30832494
1
CORR
.02920402
.24444337
.23757881
1
END DATA.
regression matrix=in(*) /dep ses/enter female race.
From: Deon De Bruin [[Link]
Sent: Friday, May 28, 2004 7:34 AM
To: Wuensch, Karl L
Subject: Correlation matrix in SPSS
Dear Dr. Wuensch
I hope you can help. After performing a first order factor analysis with oblique rotation I wish
to subject the correlations between the factors to a second order factor analysis. Can you
please give advice on how to use the correlation matrix as input for a factor analysis?
Sincerely
Deon de Bruin
From: T. Z. Keith [[Link]
Sent: Wednesday, June 02, 2004 4:17 PM
To: Bolen, Larry M
Subject: RE: Correlation matrix in SPSS
Yes, quite easy. This assumes the matrix is in an open spreadsheet. You can also read the matrix in as
part of the job. Ill see if I can find an example of that if youre interested.
Tim
_________________________________________________________
factor
/matrix=in(cor=*)
/analysis M_reas to recog
/PRINT def, corr
/plot=rotation
/CRITERIA factors(2) ITERATE(50)
/EXTRACTION paf
/ROTATION norotate
/METHOD=CORRELATION .
factor
/matrix=in(cor=*)
/analysis M_reas to recog
/PRINT def, corr
/plot=rotation
/CRITERIA factors(2) ITERATE(50)
/EXTRACTION paf
/ROTATION varimax
/METHOD=CORRELATION .
REGRESSION
matrix=in(*)
/var=M_reas to recog
/STATISTICS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT m_reas
/METHOD=ENTER compute1 compute2 r_comp1 r_comp2 recog .