Date:         Wed, 29 Sep 2010 08:44:07 +0200
Reply-To:     Marta García-Granero <mgarciagranero@gmail.com>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Marta García-Granero <mgarciagranero@gmail.com>
Subject:      Re: Frequency table when one does not have CTABLES
In-Reply-To:  <4CA2C7B7.90904@googlemail.com>
Content-Type: multipart/alternative;

  El 29/09/2010 6:59, Mark Webb escribió:
> Is it possible to product a summary frequency table [shown below]
> without CTABLES in SPSS ? The percentages shown could also be
> frequency counts.
>
>
>> Consumption
>>
>>
>>
>> Brand A
>>
>>
>>
>> Brand B
>>
>> Heavy user
>>
>>
>>
>> 30%
>>
>>
>>
>> 10%
>>
>> Medium user
>>
>>
>>
>> 60%
>>
>>
>>
>> 30%
>>
>> Light user
>>
>>
>>
>> 10%
>>
>>
>>
>> 60%
>>
>>
Hi Mark:

Maybe this is too simple, but if I understand your question correctly,
CROSSTABS would do something very close to what you want (although there
are some row&column totals you should get rid of later, perhaps using OMS):

* Fake data *.
DATA LIST LIST/consumption brand frequency (3 F8).
BEGIN DATA
1 1 3
1 2 1
2 1 6
2 2 3
3 1 1
3 2 6
END DATA.
WEIGHT BY frequency.

CROSSTABS
   /TABLES=consumption  BY brand
   /FORMAT= AVALUE TABLES
   /CELLS= COLUMN.


Best regards,
Marta GG

--
For miscellaneous SPSS related statistical stuff, visit:
http://gjyp.nl/marta/


[text/html]