Hello Everyone
I am calculating Fleiss Kappa, I have 28 raters, 5 Subjects and 5 ratings
(i.e. A dataset with 28 columns and 5 rows). The problem is that there are 2
missing values in the data. (Right now I have manually deleted the values in
csv file and then imported it in R
1) Would it better to replace those with "0" or should those be
omitted? By
omission I will be left with only 26 raters.
2) Also when we use na.omit(objectname) it deleted missing values row-wise,
how can data be omitted column-wise?
3) My second problem is that overall agreement comes to zero, whereas the
data is not showing agreement to be close to zero. The ratings 4 and 5 are
being given maximum time (the ratings are 1: total disagreement and 5: Total
Agreement) whereas for 1 and 2 its very less.
Results I got using agree()> agree(demoA)
Percentage agreement (Tolerance=0)
Subjects = 5
Raters = 26
%-agree = 0
can any1 point where I might be wrong?
4) Can we calculate percentage of agreement for each subject? If yes how?
Please help me for these queries
Thank you very much in advance
Regards
Sunita
--
View this message in context:
http://www.nabble.com/Fliess-Kappa-tp24460901p24460901.html
Sent from the R help mailing list archive at Nabble.com.