Hi everybody,
I have a question regarding coding comparison queries. I find it hard to make sense of percentage agreement. Basically, I've run a coding comparison query to see whether individual codes are reliable. Since NVivo includes negative agreement in the calculation, it is hard for me to judge whether we have achieved an acceptable percentage for the particular code. I know that there is the column "A+ B (%)" , but because this percentage is in relation to the overall source size it is not helpful either.
Do you have any suggestions for how I I can calculate positive agreement for individual codes?
I hope my question is understandable
Thanks in advance,
Marie