Limitation of κ to assess the reliability of 2 raters' judgments with different prevalences in the 2 categories
| Positive | Rater 1 Negative | Total (%) | ||
|---|---|---|---|---|
| Situation (a) | Positive | 85 | 5 | 90 |
| Rater 2 | Negative | 5 | 5 | 10 |
| κ = 0.44 | Total | 90 | 10 | 100 |
| Situation (b) | Positive | 45 | 5 | 50 |
| Rater 2 | Negative | 5 | 45 | 50 |
| κ = 0.80 | Total | 50 | 50 | 100 |