DiagnosticAgreement.deVet: Calculating agreement for dichotomous variables.

In BavoDC/AGREL: Package for calculating agreement/reliability indices

Description

Function to calculate specific agreement and overall proportion of agreement for dichotomous variables.

Usage

DiagnosticAgreement.deVet(ratings, CI = T, ConfLevel = 0.95,
  correction = c("continuity", "Fleiss", "bootstrap"), NrBoot = 1000,
  Parallel = F, no_cores = detectCores() - 1)

Arguments

ratings	A dataframe or matrix of N x P with N the number of observations and P the number of raters.
CI	Logical, indicates if confidence intervals have to be calculated
ConfLevel	The confidence level to be used in calculating the confidence intervals. Possible values are "continuity", "Fleiss" and "bootstrap".
correction	Method of calculating the confidence intervals (de Vet et al., 2017). The confidence intervals (CI) can be calculated using continuity correction, Fleiss correction or by use of bootstrap samples.
NrBoot	In case of bootstrap methodology to calculate the CI, the number of bootstrap samples.
Parallel	Logical, indicates if parallel computing has to be used to compute the confidence intervals. Implemented only when using bootstrapping to calculate the confidence intervals.
no_cores	Number of cores if parallel computing is used. Default is 1 core less than the number of cores present.

Details

This function is based on the functions as given in the appendix of the article of de Vet et al. (2017).

Value

A list with the following components:

SumTable	The summed table as defined in the article of de Vet et al. (2017)
ObservedAgreem	The overall proportion of agreement (with CI).
SpecificAgreem	The specific agreement for each of the categories (with CIs).

References

De Vet HCW, Mokkink LB, Terwee CB, Hoekstra OS, Knol DL. Clinicians are right not to like Cohen’s k. BMJ 2013;346:f2125.

de Vet, H.C.W., Terwee C.B., Knol, D.L., Bouter, L.M. (2006). When to use agreement versus reliability measures. Journal of Clinical Epidemiology, Vol.59(10), pp.1033-1039

de Vet, H.C.W., Dikmans, R.E., Eekhout, I. (2017). Specific agreement on dichotomous outcomes can be calculated for more than two raters. Journal of Clinical Epidemiology, Vol.83, pp.85-89

Examples

# Load data
data(Agreement_deVetArticle)
Df = Agreement_deVetArticle

DiagnosticAgreement.deVet(Df)

BavoDC/AGREL documentation built on May 6, 2019, 7:22 a.m.