Date:         Wed, 26 Apr 2006 16:08:43 +0930
Reply-To:     Kylie Lange <kylie.lange@flinders.edu.au>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Kylie Lange <kylie.lange@flinders.edu.au>
Subject:      Re: Crosstabs/chi square from frequency data
Comments: To: "Fredric E. Rose, Ph.D." <frose@palomar.edu>
In-Reply-To:  <C0745DB9.5C67%frose@palomar.edu>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Hi Frederick,

Yes you're right, there is a way to get SPSS to work with the summary
numbers you have. Create a data file with three variables, one for the
frequency counts, one to indicate the condition, and one to indicate
whether they agreed. Ie:

cnt cond  agr
62    A    Y
38    A    N
48    B    Y
52    B    N

Then, weight your data set by the 'cnt' variable (Data menu > Weight
cases). SPSS will now treat your data file as if it had all 200 rows.
You can then go into crosstabs and create a table of 'cond' versus 'agr'
and request the chi-square test as per normal.

Hope this helps.

Cheers,
Kylie.


On 26/04/2006 3:44 PM, Fredric E. Rose, Ph.D. wrote:
> OK, I know calculating a chi square by hand is about the easiest thing known
> to human kind, but I was wondering if it is possible to get SPSS to do it
> from already-calculated frequency data rather than raw data?
>
> For example, in condition A 62 individuals agreed to a request and 38 did
> not.  In condition B, 48 individuals agreed and 52 did not.  So, can SPSS do
> the calculations, etc., based on those numbers, without having the responses
> for 200 individuals?  I know I could create the 200 data points, but this is
> more of a curiosity than anything.
>
> TIA,
>
> F
> --
> Fredric E. Rose, Ph.D.
> Assistant Professor of Psychology
> Palomar College
> (760) 744-1150 x2344
> frose@palomar.edu