Answers are grounded in this paper and cite their sources.
Based on the different formulae used to calculate the level of chance-corrected agreement, Gwet's AC1 was shown to provide a more stable inter-rater reliability coefficient than Cohen's Kappa. It was also found to be less affected by prevalence and marginal probability than that of Cohen's Kappa, and therefore should be considered for use with inter-rater reliability analysis.
No takes yet. Share an insight, caveat, or question.
Wongpakaran et al. (2013) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: