Average in crosstabs

[wollyka]

Hello
In a survey we are doing, we are asking the top three countries the
respondent export to and the percent of export for each country from
the total export.. In SPSS, we have then 3 Variables for countries:
one for each country and another three variables for the percentage..
The first variable for the country is obviously paired with the first
variable for export percentage..

Now to know the percentage of export for each country for the total
sample , we do a paired crosstab.. However,this is not exactly what i
want.. i want to calculate for each country mentioned by the
respondents the average percentage of export to it.. how can it be
done in SPSS? i usuaaly copy it to Excel and the do the calculations
manually, is there a quicker way to do it in SPSS?

Thanks for the help

[wollyka]

to clarify more i need to create a two columns table one column with
the countries and the other with the average percent of exports to
this country
thanks

[wollyka]

no one? :(

[Bruce Weaver]

If I follow, your data file looks something like this:

RespID C1 C2 C3 P1 P2 P3

where C1-C3 hold the top 3 country names for that ID, and P1-P3
hold the percentage of exports to those countries.

If this is correct, I think I would frestructure the file (with
VARSTOCASES) to a format with 3 rows per respondent, and these
variables:

RespID C P
1 1 %1
1 2 %2
1 3 %3
2 1 %1
2 2 %2
2 3 %3
etc

Now you have all the countries in a single column, and should be
able to easily get the table you want.

If the country variables are strings, be watchful for variations
on how the same country is entered.

--
Bruce Weaver
bwe...@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/
"When all else fails, RTFM."

[wollyka]

> bwea...@lakeheadu.cahttp://sites.google.com/a/lakeheadu.ca/bweaver/

> "When all else fails, RTFM."

Thanks for the help, i managed to get all all the countries in one
column, then i used split data (organize by countries) and then ran
frequencies on the percentages to obtain the mean of percentage for
each country.. it works but is it possible to have it in one table as
the below example:
Country Percentage average
C1 10%
C2 24%
C3 33%
etc..
Thanks again

[Bruce Weaver]

wollyka wrote:

>
> Thanks for the help, i managed to get all all the countries in one
> column, then i used split data (organize by countries) and then ran
> frequencies on the percentages to obtain the mean of percentage for
> each country.. it works but is it possible to have it in one table as
> the below example:
> Country Percentage average
> C1 10%
> C2 24%
> C3 33%
> etc..
> Thanks again

MEANS or DESCRIPTIVES should give you that type of output. E.g.,
if P and C are your Percentage and Country variables,

means P by C.

You can add a subcommand (/CELLS = mean, I think--no SPSS on this
machine) to limit the output to the means, if that's all you want.

The SUMMARIZE command can produce output like that too.

--
Bruce Weaver
bwe...@lakeheadu.ca

[wollyka]

On Apr 5, 10:08 pm, Bruce Weaver <bwea...@lakeheadu.ca> wrote:
> wollyka wrote:
>
> > Thanks for the help, i managed to get all all the countries in one
> > column, then i used split data (organize by countries) and then ran
> > frequencies on the percentages to obtain the mean of percentage for
> > each country.. it works but is it possible to have it in one table as
> > the below example:
> >  Country  Percentage average
> >    C1            10%
> >    C2             24%
> >    C3             33%
> > etc..
> > Thanks again
>
> MEANS or DESCRIPTIVES should give you that type of output.  E.g.,
> if P and C are your Percentage and Country variables,
>
> means P by C.
>
> You can add a subcommand (/CELLS = mean, I think--no SPSS on this
> machine) to limit the output to the means, if that's all you want.
>
> The SUMMARIZE command can produce output like that too.
>
> --
> Bruce Weaver

> bwea...@lakeheadu.cahttp://sites.google.com/a/lakeheadu.ca/bweaver/

> "When all else fails, RTFM."

Thanks summarize worked like a charm :)