I am calculating an inter-rater agreement for a biological question. Basicaly, 6 readers are evaluated on 60 subjects with 2 possibilities categorical scale for each subjects. I am using kappam.fleiss function on R to do so.
My question is: how to take into account missing data ? If I put "NA" on my database, I am scared that NA would be interpreted as a new entry for the categorical interpretation. and thus would be associated with an abnormal kappa value.
Thanks a lot.
Thomas
You could define during import, which cells should be recognized as NAs. See more info here. When you have NAs in the data, R is able to treat it as missing and not as valu. See more here.