For this I am wondering if you have a set of qualitative responses to questions (e.g., how would you define the term 'mental health'?) and have developed a coding scheme what level of inter-rater reliability would you be looking for in order to indicate that the coding system is reliable? Note this is just an example of a question but there are many more we could have.
David L Morgan
I have recommendations for .80, but this is basically a matter of judgement where there is no specific standard.
0 votes
0 thanks
Meiko Makita
It would also depend on what you are coding. We often code ‘sentiment’ on tweets communication and since it’s so subjective we do get lower percentages than with the rest of the variables. We use this Cohen kappa guide (photo attached).
1 votes
0 thanks
- Badges
- Science topic
- Similar topics
- Statistics
- Survey
More Meredith R Wilkinson's questions See All
Similar questions and discussions