[RkWard-devel] Ordinal nominal

[Thomas Friedrichsmeier]

Hi Ilias,

On Thursday 29 March 2007 05:55, I. Soumpasis wrote:
> However there were some ordinal data which SPSS has a way to handle them.
> For example for variable ABC in the variables tab you declare:
> from 1 to 4
> 1 as A
> 2 as B
> 3 as C
> 4 as D
> Thus it can run the wilcoxon tests and other no parametric needed. But when
> I import the data the variables are imported as factor variables and the
> wilcoxon tests don't play.
>
> >>From your knowledge in R is there a way we can handle this? Can we do
>
> something with RKWard?

Well, you can convert it using
	as.numeric(x)
. Wrapping something like this around the variables passed to the test in 
RKWard would not be too hard, of course. The tricky thing is to find out, 
when this is needed, and how it could sensibly be controlled in the GUI (it 
probably should not be done unconditionally, as it's a sensible safeguard 
against feeding the wrong kind of data into the test). Perhaps we could add 
an option "Coerce data to numeric"? This would likely be needed in a whole 
bunch of tests, though, so we should think for a while, whether this is 
really the most elegant solution available. Comments?

Regards
Thomas
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://mail.kde.org/pipermail/rkward-devel/attachments/20070329/f8aa0ebd/attachment.sig>