|
|
Here is an example using the cars.sav
file shipped with SPSS. Explanation below.
temporary.
if missing(origin) omissing = 1.
value labels omissing 1 'Missing'.
variable label omissing ' '.
CTABLES
/TABLE year BY origin [COUNT
ROWPCT.COUNT] + omissing [C][COUNT]
/SLABELS POSITION=ROW
/CATEGORIES VARIABLES=year origin
ORDER=A KEY=VALUE EMPTY=INCLUDE MISSING=EXCLUDE
/CATEGORIES VARIABLES=omissing
[1] .
omissing is a temporary variable that
will be 1 if the column variable, origin, is missing.
It has a blank variable label in order
to suppress the variable name in the output.
SLABELS moves the statistics to the
rows.
By concatenating omissing in the columns,
you get an extra subtable that tabulates the missing values while they
are excluded from the main body of the table. Doing it this way allows
the use of COUNT for both variables, so no extra labeling row is required
for the statistics.
A number of variations are possible,
but this should give you the idea.
HTH,
Jon Peck
Jon Peck
SPSS, an IBM Company
[hidden email]
312-651-3435
Dear list,
I would liketo create a ctables omits a user-defined
missing (a NA category) from the denominator for row percentages, but report
the raw number of NA's at the end of the row.
I am looking for something like this abbreviated
example:
Satisfied
Dissatisfied
Total
NA
Count Row
% Count
Row % Count Row %
Count
Pricing
7 70
3 30
10 100
3
While I'm at it, is there a way to print
the row % under the count, instead of beside it?
Thank you,
John
|
Locked
