I used Kappa for our PT school research project. I just reviewed it
yesterday for other purpose. (I am currently taking a baby statistics
course.)
What I can find is:
z = kappa/square root of variance of kappa.
I have never seen z = k/standard error of Kappa before.
But I guess the only diffence between these two would be consideration of
sample size. The larger the sample size, the smaller the standard error
and therefore the z score above will become larger.
square root of variance of kappa = standard error x square root of sample
size
(So, if there is only one sample, then both are identical)
Does it make sense to you? My knowledge of statistics is very limited
and I hope that I did not confuse you even further.
Sachiko Komagata
"Geert.Aufdemkampe" wrote:
> Dear all,
>
> I am a bit confused. There seems to be two ways to calculate the
> significance of Kappa.
> Siegel and Castellan propose to to divide Kappa by the square root of
> the variance of Kappa, whereas Haas proposes to divide Kappa by
> the standard error of Kappa.
>
> Can anyone help?
>
> Geert
>
> Geert Aufdemkampe MScPT, senior research associate
> Polytechnic of Utrecht, Faculty of Health Science,
> Department of Physical Therapy, Utrecht, The Netherlands
> phone: ..31302585136 fax: ..31302540608
> e-mail: [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|