PRINT Subcommand (MULTIPLE CORRESPONDENCE command)
The Model Summary statistics (Cronbach's alpha
and the variance accounted for) and the HISTORY
statistics (the variance accounted for, the loss, and the increase
in variance accounted for) for the last iteration are always displayed.
That is, they cannot be controlled by the PRINT
subcommand. The PRINT
subcommand controls the display of optional additional output.
The output of the MULTIPLE CORRESPONDENCE
procedure is always based on the transformed variables. However,
the correlations of the original variables can be requested as well
by the keyword OCORR
.
The default keywords are DESCRIP
, DISCRIM
, and CORR
. That is, the three keywords are in
effect when the PRINT
subcommand
is omitted or when the PRINT
subcommand
is given without any keywords. Note that when some keywords are specified,
the default is nullified and only the keywords specified become in
effect. If a keyword that cannot be followed by a varlist
is duplicated or if a contradicting
keyword is encountered, then the later one silently becomes effective
(in case of a contradicting use of NONE
, only the keywords following NONE
are effective). For example,
/PRINT <=> /PRINT = DESCRIP DISCRIM CORR
/PRINT = DISCRIM DISCRIM <=> /PRINT = DISCRIM
/PRINT = DISCRIM NONE CORR <=> /PRINT = CORR
If a keyword that can be followed by a varlist
is duplicated, it will cause a syntax
error and the procedure will terminate. For example, /PRINT = QUANT QUANT
is a syntax error.
The following keywords can be specified:
DESCRIP(varlist). Descriptive
statistics (frequencies, missing values, and mode). The
variables in the varlist
must
be specified on the VARIABLES
subcommand, but need not appear on the ANALYSIS
subcommand. If DESCRIP
is not followed by a varlist
, Descriptives tables are displayed for all the variables in the varlist
on the ANALYSIS
subcommand.
DISCRIM. Discrimination measures per variable and per dimension.
QUANT(varlist). Category quantifications
(centroid coordinates), mass, inertia of the categories, contribution
of the categories to the inertia of the dimensions, and contribution
of the dimensions to the inertia of the categories. Any
variable in the ANALYSIS
subcommand
may be specified in parentheses after QUANT
. If QUANT
is not followed by
a varlist
, Quantification tables
are displayed for all variables in the varlist
on the ANALYSIS
subcommand. Note: the coordinates and contributions
(including the mass and inertia) are displayed in separate layers
of the pivot table output, with the coordinates shown by default.
To display the contributions, activate (double-click) on the table
and select Contributions from the Layer dropdown list.
HISTORY. History of iterations. For each iteration, the variance accounted for, the loss, and the increase in variance accounted for are shown.
CORR. Correlations
of the transformed variables, and the eigenvalues of this correlation
matrix. Correlation tables are displayed for each set
of quantifications, thus there are ndim (the number of dimensions in the analysis) correlation tables; the ith table contains the correlations of the
quantifications of dimension i, i = 1, ..., ndim. For variables with missing values
specified to be treated as PASSIVE
on the MISSING
subcommand,
the missing values are imputed according to the specification on the
PASSIVE
keyword (if nothing
is specified, mode imputation is used).
OCORR. Correlations
of the original variables, and the eigenvalues of this correlation
matrix. For variables with missing values specified to
be treated as PASSIVE
on the
MISSING
subcommand, the missing
values are imputed with the variable mode.
OBJECT((varname)varlist). Object scores
(component scores) and, in separate table, mass, inertia of the objects,
contribution of the objects to the inertia of the dimensions, and
contribution of the dimensions to the inertia of the objects. Following the keyword, a varlist
can be given in parentheses to display variables (category indicators)
along with the object scores. If you want to use a variable to label
the objects, this variable must occur in parenthesis as the first
variable in the varlist
. If no
labeling variable is specified, the objects are labeled with case
numbers. The variables to display along with the object scores and
the variable to label the objects must be specified on the VARIABLES
subcommand but need not appear
on the ANALYSIS
subcommand. If
no varlist
is given, only the
object scores are displayed.
NONE. No optional
output is displayed. The only output shown is the Model
Summary and the HISTORY
statistics
for the last iteration.
The keyword TO
in a variable list can only be used with variables that are in the ANALYSIS
subcommand, and TO
applies only to the order of the variables
in the ANALYSIS
subcommand.
For variables that are in the VARIABLES
subcommand but not in the ANALYSIS
subcommand, the keyword TO
cannot
be used. For example, if /VARIABLES = v1
TO v5
and the ANALYSIS
subcommand has /ANALYSIS v2 v1 v4
, then /PLOT OBJECT(v1 TO v4)
will give two object plots, one labeled with v1 and one labeled with v4. (/PLOT OBJECT(v1 TO v4 v2 v3 v5)
will give objects plots labeled with v1, v2, v3, v4, and v5).