DOI: 10.1007/S11135-012-9745-9
关键词: Inter-rater reliability 、 Statistics 、 Psychology 、 Monte Carlo method 、 Response Parameters 、 Reliability (statistics) 、 Practical implications 、 Econometrics
摘要: Although it has long been a consensus that intercoder reliability is crucial to the validity of content analysis study, choice among them debated. This study reviewed and empirically tested most popular indices, aiming find robust index against prevalence rater bias, by testing their relationships with response surface methodology through Monte Carlo experiment. It was found Maxwell’s R.E superior Krippendorff’s α, Scott’s π, Cohen’s κ, I r Perreault Leigh, Gwet’s AC 1. More nuanced prevalence, sensitivity, specificity indices were discovered plots. Both theoretical practical implications also discussed in end.