PRINT Subcommand (FACTOR command)
PRINT
controls
the statistical display in the output.
- Keywords
INITIAL
,EXTRACTION
, andROTATION
are the defaults ifPRINT
is omitted or specified without keywords. - If any keywords are specified, only the output specifically requested is produced.
- The requested statistics are displayed only for variables
specified on the last
ANALYSIS
subcommand. - If more than one
PRINT
subcommand is specified, the last is in effect. - If any
ANALYSIS
subcommand is explicitly specified, allPRINT
subcommands specified before the lastANALYSIS
subcommand are ignored. If noPRINT
subcommand is specified after the lastANALYSIS
subcommand, the default takes effect.
INITIAL. Initial communalities for each variable, eigenvalues of the unreduced correlation matrix, and percentage of variance for each factor.
EXTRACTION. Factor pattern matrix, revised communalities, the eigenvalue of each factor retained, and the percentage of variance each eigenvalue represents.
ROTATION. Rotated factor pattern matrix, factor transformation matrix, factor correlation matrix, and the post-rotation sums of squared loadings.
UNIVARIATE. Valid number
of cases, means, and standard deviations. (Not available
with matrix input.) If MISSING=MEANSUB
or PAIRWISE
, the output also
includes the number of missing cases.
CORRELATION. Correlation matrix. Ignored if the input is a factor-loading matrix.
COVARIANCE. Covariance
matrix. Ignored if the input is a factor-loading matrix
or a correlation matrix that does not contain standard deviations
(STDDEV
or SD
).
SIG. Matrix of significance levels of correlations.
DET. Determinant
of the correlation or covariance matrix, depending on the specification
on METHOD
.
INV. Inverse of
the correlation or covariance matrix, depending on the specification
on METHOD
.
AIC. Anti-image covariance and correlation matrices 1. The measure of sampling adequacy for the individual variable is displayed on the diagonal of the anti-image correlation matrix.
KMO. Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett’s test of sphericity. Always based on the correlation matrix. Not computed for an input matrix when it does not contain N values.
REPR. Reproduced
correlations and residuals or reproduced covariance and residuals,
depending on the specification on METHOD
.
FSCORE. Factor score
coefficient matrix. Factor score coefficients are calculated
using the method requested on the SAVE
subcommand. The default is the regression method.
ALL. All available statistics.
DEFAULT. INITIAL
, EXTRACTION
, and ROTATION
.
Example
FACTOR VARS=V1 TO V12
/SELECT=COMPLETE (‘yes’)
/MISS=MEANSUB
/PRINT=DEF AIC KMO REPR
/EXTRACT=ULS
/ROTATE=VARIMAX.
- This example specifies a factor analysis that includes all variables between and including V1 and V12 in the active dataset.
- Only cases with the value “yes” on COMPLETE are used.
- Variable means are substituted for missing values. Only values for the selected cases are used in computing the mean. This mean is used to substitute missing values in analyzing the selected cases and in computing factor scores for all cases.
- The output includes the anti-image correlation and covariance matrices, the Kaiser-Meyer-Olkin measure of sampling adequacy, the reproduced correlation and residual matrix, as well as the default statistics.
- Factors are extracted using unweighted least squares.
- The factor pattern matrix is rotated using the varimax rotation.