Abstract
The authors start their review with the discussion on the kappa statistic commonly used in agreement studies. The practical appeal of single summary statistics to summarize an entire statistical analysis such as the ubiquitous R2 cannot be denied. However, one should always be quite cautious in concluding solely based on these statistics. As indicated by the authors, the kappa statistic is no different. Often defined as a sample statistic, potentially lead to inconsistent results. In his early work in this area, Professor de Mast explored the paradoxical behavior of this statistic. His more recent work on introducing more sophisticated methods in alleviating some of the issues discussed in this article is certainly worth praise. However, considering the widespread use of the kappa statistic, it is also worth looking into some of the authors' comments on the behavior of this statistic.
Original language | English |
---|---|
Journal | Quality Engineering |
Volume | 26 |
Issue number | 1 |
Pages (from-to) | 40-43 |
ISSN | 0898-2112 |
Publication status | Published - 2014 |