Confidence Interval Approach for the Number of Subjects Required

Power Analysis for Interobserver Agreement Analysis.

Usage

kappaSizeCI(
  outcome = "2",
  citype = "two_sided",
  kappa0 = 0.6,
  kappaL = 0.4,
  kappaU = 0.8,
  props = "0.20 , 0.80",
  raters = "2",
  alpha = 0.05
)

Arguments

outcome: Number of outcome level.
citype: Type of confidence interval: 'two_sided' or 'one_sided'.
kappa0: The preliminary (anticipated) value of kappa - the agreement you expect to observe, around which the confidence interval is planned. kappaSize documents this argument as "the preliminary value of kappa". It is NOT a null hypothesis value; contrast kappaSizePower, where kappa0 IS the null being tested.
kappaL: The lower limit of the kappa.
kappaU: The upper limit of the kappa.
props: Proportions of outcome level.
raters: Number of raters.
alpha: The significance level.

Value

A results object containing:

`results$notices`					a html
`results$text1`					a preformatted
`results$text_summary`					a preformatted
`results$text2`					a preformatted