caseids instead of counts in cross tab

[s_pushparaj]

Dear List members
I have a problem related to getting the caseids in the crosstab instead of the counts in them. Is it possible to get them.
kindly help
with advance thanks
pushparaj

[ViAnn Beadle]

Why do you want to do this and what do you expect to happen if you have
multiple cases within a cell?

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
s_pushparaj
Sent: Monday, August 27, 2007 12:01 PM
To: [hidden email]
Subject: caseids instead of counts in cross tab

Dear List members
I have a problem related to getting the caseids in the crosstab instead of
the counts in them. Is it possible to get them.
kindly help
with advance thanks
pushparaj

[Hector Maletta]

         If that if what you want, you should use the CASEID variable as one
of the variables in the crosstab. For instance: CROSSTABS CASEID BY SEX will
produce a listing of all cases (one per row) classified in two columns, one
for each sex, with a frequency of 1 in the corresponding cell of each row.
CROSSTABS CASEID BY REGION BY SEX will produce a table where all members of
one gender appear first (one person per row), and all members of the
opposite gender appear below, with one column per region, and a frequency of
1 in the cell corresponding to the region where each person lives.
         Now, this would be a rather foolish pair of tables to produce. It
is more like a listing than a statistical table. You may want to produce
such an individualized case-by-case report, but not for statistical
analysis.
         Think again.

         Hector

         -----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
s_pushparaj
Sent: 27 August 2007 15:01
To: [hidden email]
Subject: caseids instead of counts in cross tab

         Dear List members
         I have a problem related to getting the caseids in the crosstab
instead of the counts in them. Is it possible to get them.
         kindly help
         with advance thanks
         pushparaj

[Richard Ristow]

At 02:01 PM 8/27/2007, s_pushparaj wrote:

>I have a problem related to getting the caseids in the crosstab
>instead of the counts in them. Is it possible to get them.

You've had some correct answers, including questions whether this is a
good idea.

At 02:38 PM 8/27/2007, Hector Maletta wrote:

>   Now, this is more like a listing than a statistical table.

which it is. Look at the LIST command, possibly with subcommand

  /FORMAAT=NUMBERED

[Peck, Jon]

I can see how this could be quite useful: a crosstab layout with a list of case id's replacing the usual count and other statistics.  It would give you a nice visual picture of the distribution as long as the number of cases and categories is small enough for the display to be clear.  Of course, this would not be for inferential purposes.

But as a visual, you could easily get this sort of display by doing a scatterplot of the x and y categorical variables.  To see all the points, if you use the Chart Builder, click the checkbox "stack identical values", and to identify the cases, turn on data value labels in the Chart Editor.

Again, this will only be useful if the number of cases and categories is fairly small, but I doubt that you would want huge numbers of id's anyway.  And you can enlarge the chart in order to see more points and ids.

HTH,
Jon Peck

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Richard Ristow
Sent: Monday, August 27, 2007 2:37 PM
To: [hidden email]
Subject: Re: [SPSSX-L] caseids instead of counts in cross tab

At 02:01 PM 8/27/2007, s_pushparaj wrote:

>I have a problem related to getting the caseids in the crosstab
>instead of the counts in them. Is it possible to get them.

You've had some correct answers, including questions whether this is a
good idea.

At 02:38 PM 8/27/2007, Hector Maletta wrote:

>   Now, this is more like a listing than a statistical table.

which it is. Look at the LIST command, possibly with subcommand

  /FORMAAT=NUMBERED