ttest.fleiss: Paired t-test for the difference of 2 Fleiss generalized...

In kgwet/pairedCAC: Testing the difference of correlated agreement coefficients for statistical significance

Description

This function performs the paired t-test for testing the difference between two correlated Fleiss kappa coefficients for statistical significance. It implements the linearization method of Gwet (2016).

Usage

ttest.fleiss(
  g1.ratings,
  g2.ratings,
  weights = "unweighted",
  conflev = 0.95,
  N = Inf
)

Arguments

g1.ratings	is a mandatory parameter representing the first data frame of ratings.
g2.ratings	is a mandatory parameter representing the second data frame of ratings.
weights	is an optional parameter that defines the weights needed in a weighted analysis. It's default value is “unweighted” which requests the unweighted analysis.
conflev,	is an optional parameter representing the confidence level. It's default value is 0.95.
N	is an optional parameter representing the size of the subject population. It's default value is infinity.

Details

This function takes 2 required parameters, which are the 2 groups of raters being compared: "g1.ratings" and "g2.ratings." Both datasets must have the exact same number of rows, and each column represents one rater and contains its ratings (numeric or alphabetic). The user must exclude all subjects that are not rated by any rater.

References

Fleiss, J.L. (1971). “Measuring nominal scale agreement among many raters,” Psychological Bulletin, 76, 378-382.

Gwet, K.L. (2016). “Testing the Difference of Correlated Agreement Coefficients for Statistical Significance, Educational and Psychological Measurement, 76(4) 609-637.

kgwet/pairedCAC documentation built on Dec. 21, 2021, 6:37 a.m.