Finding Kappa from a List of Outcomes
Sometimes you only have a list of outcomes. For instance, say I had two raters assign people to one of five categories. All that I know is the frequency of occurrence for each type of rating. In other words, I know that both raters chose 1 once, both raters chose 2 twice, one rater chose 1 and the other chose 2 twice, etc.
In this case, you can have Stata make a table and commit that table to memory, effectively entering the data for you. The command is tabi row of outcomes \ row of outcomes \ row of outcomes \ etc, replace where you are building an outcome matrix using rows of outcome frequencies. The replace option replaces the existing data set with the frequency data.
Entering the data from the two-rater example as a table would look like the following tabi 1 2 0 0 0 \ 0 2 1 0 0 \ 0 2 1 0 0\ 0 0 0 2 0 \ 0 0 1 0 1, replace and yield the following output (which you can ignore after checking for any errors you may have made during input).
Then, if you browse your data, you will see it is now in the perfect form to use kappa! The command to run kappa on your new data is kap row col [freq=pop] with the square brackets.