The package also includes tools to assess agreement between raters and to plan reliability studies.
Example Dataset
df_agr <- read.csv(system.file("extdata", "agreement_example.csv", package = "meddecide"))
head(df_agr)Calculating Agreement
Use the agreement() function to compute Cohen’s or
Fleiss’ Kappa statistics depending on the number of raters.
agr_res <- agreement(data = df_agr)
agr_res$kappaSample Size for Agreement Studies
The functions kappaSizeCI(),
kappaSizeFixedN() and kappaSizePower() help
determine the sample size needed for agreement research.
# precision based approach
kappaSizeCI(kappa0 = 0.7, conf.level = 0.95, w = 0.1)
# fixed number of raters
kappaSizeFixedN(kappa0 = 0.7, n = 60)
# power based approach
kappaSizePower(kappa0 = 0.7, kappa1 = 0.8, power = 0.8)These calculations support planning of reliability experiments in clinical research.