FREQUENCIES-Question: How to display a category with no respondants in it?

[Walter Funk]

Hi,

I know I've read about this, but can't remember were to find it:

We have a variable with some valid answer categories/values not having
been ticked by any respondant. For example, if the valid categories are
1, 2, 3, 4 and 5  - we have only respondants answering 1, 2, 4 and 5.

But we want to have a Frequency-table with all valid answer categories
being displayed, that is: also showing the value "3". How can we force
SPSS to display all valid categories of a variable no matter whether
respondants chose that answer or not?

Thanks for your help
Walter

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Art Kendall]

  One way is to add a dummy case with a tiny weight.
Another is to use custom tables.

data list list/id (n2) a b c d e (5f1).
begin data
01 1 2 3 4 5
02 5 4 3 2 1
03 2 3 4 5 1
04 1 4 5 1 2
05 4 5 1 2 2
06 5 1 2 3 4
07 3 3 3 3 3
end data.
formats id (n2).
value labels a to e
  1 'strongly disagree'
  2 'disagree'
  3 'neither'
  4 'agree'
  5 'strongly agree'.
compute wgt = 1.
if id eq 7 wgt = 1e-32.
weight by wgt.
frequencies vars = a to e.
select if id ne 7.
weight off.
* Custom Tables.
CTABLES
   /VLABELS VARIABLES=a b c d e DISPLAY=BOTH
   /TABLE a [COUNT F40.0] + b [COUNT F40.0] + c [COUNT F40.0] + d [COUNT
F40.0] + e [COUNT F40.0]
   /CLABELS ROWLABELS=OPPOSITE
   /CATEGORIES VARIABLES=a b c d e ORDER=A KEY=VALUE EMPTY=INCLUDE.
* Custom Tables.
CTABLES
   /VLABELS VARIABLES=a e DISPLAY=BOTH
   /TABLE a [COUNT F40.0] + e [COUNT F40.0]
   /CLABELS ROWLABELS=OPPOSITE
   /CATEGORIES VARIABLES=a e ORDER=A KEY=VALUE EMPTY=INCLUDE.

Art Kendall
Social Research Consultants


On 10/13/2010 7:32 AM, Walter Funk wrote:

> Hi,
>
> I know I've read about this, but can't remember were to find it:
>
> We have a variable with some valid answer categories/values not having
> been ticked by any respondant. For example, if the valid categories are
> 1, 2, 3, 4 and 5  - we have only respondants answering 1, 2, 4 and 5.
>
> But we want to have a Frequency-table with all valid answer categories
> being displayed, that is: also showing the value "3". How can we force
> SPSS to display all valid categories of a variable no matter whether
> respondants chose that answer or not?
>
> Thanks for your help
> Walter
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
>

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Art Kendall]

If you have not already found out which value is never endorsed by any respondent and you need the output as formatted by FREQUENCIES you could do the ctables approach first and then put in a dummy case  something like this in the data
07 0 2 3 1 4
and
missing values a to e (0)
in this example the dummy case says that there was no unendorsed value for variable a. Nobody endorsed 2 for variable b. Nobody endorsed 3 for variable c.
Nobody endorsed 1 for variable d. Nobody endorsed 4 for variable e.



Art
On 10/13/2010 10:15 AM, Art Kendall wrote:

 One way is to add a dummy case with a tiny weight.
Another is to use custom tables.

data list list/id (n2) a b c d e (5f1).
begin data
01 1 2 3 4 5
02 5 4 3 2 1
03 2 3 4 5 1
04 1 4 5 1 2
05 4 5 1 2 2
06 5 1 2 3 4
07 3 3 3 3 3
end data.
formats id (n2).
value labels a to e
 1 'strongly disagree'
 2 'disagree'
 3 'neither'
 4 'agree'
 5 'strongly agree'.
compute wgt = 1.
if id eq 7 wgt = 1e-32.
weight by wgt.
frequencies vars = a to e.
select if id ne 7.
weight off.
* Custom Tables.
CTABLES
  /VLABELS VARIABLES=a b c d e DISPLAY=BOTH
  /TABLE a [COUNT F40.0] + b [COUNT F40.0] + c [COUNT F40.0] + d [COUNT
F40.0] + e [COUNT F40.0]
  /CLABELS ROWLABELS=OPPOSITE
  /CATEGORIES VARIABLES=a b c d e ORDER=A KEY=VALUE EMPTY=INCLUDE.
* Custom Tables.
CTABLES
  /VLABELS VARIABLES=a e DISPLAY=BOTH
  /TABLE a [COUNT F40.0] + e [COUNT F40.0]
  /CLABELS ROWLABELS=OPPOSITE
  /CATEGORIES VARIABLES=a e ORDER=A KEY=VALUE EMPTY=INCLUDE.

Art Kendall
Social Research Consultants


On 10/13/2010 7:32 AM, Walter Funk wrote:

Hi,

I know I've read about this, but can't remember were to find it:

We have a variable with some valid answer categories/values not having
been ticked by any respondant. For example, if the valid categories are
1, 2, 3, 4 and 5  - we have only respondants answering 1, 2, 4 and 5.

But we want to have a Frequency-table with all valid answer categories
being displayed, that is: also showing the value "3". How can we force
SPSS to display all valid categories of a variable no matter whether
respondants chose that answer or not?

Thanks for your help
Walter

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD