[rkward-devel] power analysis plugin

[Thomas Friedrichsmeier]

Hi Meik,

On Wednesday 08 October 2014 11:49:56 meik michalke wrote:
> Am Dienstag, 7. Oktober 2014, 13:43:33 schrieb Thomas Friedrichsmeier:
> > - For the two sample tests, when estimating the size of one of the
> > samples,
> > why not always make it the second sample (i.e. first sample size
> > provided)?
> 
> fixed (in the script).

hm, what I meant was: Why not get rid of that radio control, completely, not 
just changing the default. I can see that for the R function signature, it 
would have been more cumbersome to explain "Exactly one of the parameters 
needs to be NULL, except it can't be n1". But for the GUI, I don't see the 
point of showing an additional control "Now select, whether you want to 
estimate n1 given n2 or n2 given n1", when it really doesn't make any 
difference (but adds complexity both to the UI, and to the logic behind it).

Then, after that, I had the idea to get rid of the other hidden radio 
(controlling which df to estimate for GLM), too. My plan was:
1. Move test specification to the left, making it the "first step", logically
2. In the "target measure" radio, add a fifth option "numerator df", which 
would be enabled for GLM, only.
The main reason is that - logically - it makes little sense to read numerator 
df as a sub-item of sample size, and I had difficulty thinking up an 
appropriate help snippet due to this.

Well, when I wanted to make that experiment, I found that rkwarddev does not 
yet handle ids on radio-options. And then I found out, that the fact that 
radio-options can be disabled, dynamically, was not really documented, so far. 
(See Import Text / CSV Data-Plugin for an example usage). I have fixed the 
documentation. Could you add this to rk.XML.option()?

Regards
Thomas
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/rkward-devel/attachments/20141011/b679bc00/attachment.sig>