|Date: ||Fri, 26 May 2006 18:54:45 -0400|
|Reply-To: ||Keith McCormick <firstname.lastname@example.org>|
|Sender: ||"SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>|
|From: ||Keith McCormick <email@example.com>|
|Subject: ||Re: [Table layout]|
|Content-Type: ||text/plain; charset=ISO-8859-1; format=flowed|
I became especially interested in this post because I have just been
teaching Tables this week. I am convinced that in its current form,
this is not possible. You can not have scale variables in both the
row and column dimension. There is nothing in base that would let you
do this either.
You will have to transform the variables so that all the colors, etc.
are in one column with a new variable for brand next to it, or vice
versa. Once you do so, it should be pretty simple. Sorry for the
rearranging, but I hope this helps. If you don't have tables, you
might want to reply to this post to that effect, but it you rearrange
it you should be able to work it out.
On 5/25/06, David Lindsay <firstname.lastname@example.org> wrote:
> I have a bit of a brain block at the moment and some help would
> be very useful.
> I have a number of variables q32 q32 q33 q34 q35 ....q66 that
> are answers to 5 point likert scales, that refer to 5 brands
> that have been consumer tested
> and I want a to create a table of means that in layout terms
> looks like this:
> brand1 brand2 brand3 brand4 brand5
> colour q33 q40 q47 q54 q61
> shape q34 q41 q48 q55
> useful q35 q42
> Thanks for your help.