Package 'catekappa'

catekappa: Design and Analysis of Categorical Agreement Tests Based on Kappa Statistics

Description

CATEKAPPA (Categorical Agreement Test Evaluation) provides a Shiny interactive application and supporting functions for the design and analysis of categorical agreement tests.

Details

This package wraps the core functionality of the irr and kappaSize packages:

• Design stage: Use calc_sample_size_kappa to calculate sample size, supporting 2–5 categories and 2+ raters.

• Analysis stage: Use analyze_kappa to compute Cohen's, Fleiss', and Light's Kappa statistics.

• Interactive app: Use run_cate_app to launch the Shiny interface.

Author(s)

Gai Zheng [email protected] Xincheng Li [email protected] Yingjie Jiangwang [email protected] Panwei Zhao [email protected]

See Also

run_cate_app, calc_sample_size_kappa, analyze_kappa

---

Analyze Agreement Using Kappa Statistics

Description

Analyze consistency between raters using Cohen's, Fleiss', or Light's Kappa. Wraps functions from the irr package.

Usage

analyze_kappa(data, type = "cohen", detail = FALSE)

Arguments

data	A data frame or matrix with subjects as rows and raters as columns.
type	Type of kappa: "cohen" (2 raters), "fleiss" (3+ raters), or "light" (3+ raters, pairwise).
detail	Logical. If TRUE, returns detailed output including individual scores.

Value

A list with kappa results, interpretation, and data summary.

Examples

data <- data.frame(
  Rater1 = c("Yes", "No", "Yes", "Yes", "No"),
  Rater2 = c("Yes", "No", "Yes", "No", "No")
)
analyze_kappa(data, type = "cohen")

---

Calculate Sample Size for Kappa Statistic

Description

Calculate required sample size for consistency tests. Directly wraps kappaSize::PowerBinary / Power3Cats / Power4Cats / Power5Cats.

Usage

calc_sample_size_kappa(
  kappa0 = 0.4,
  kappa1 = 0.6,
  props = c(0.5, 0.5),
  alpha = 0.05,
  power = 0.8,
  raters = 2
)

Arguments

kappa0	Null hypothesis value of kappa (H0).
kappa1	Alternative hypothesis value of kappa (H1).
props	Expected proportions of categories. Must sum to 1.
alpha	Significance level. Default 0.05.
power	Desired power. Default 0.8.
raters	Number of raters (>=2). Default 2.

Value

A list with sample size n and parameters.

Examples

calc_sample_size_kappa(kappa0 = 0.4, kappa1 = 0.6, props = c(0.5, 0.5))
calc_sample_size_kappa(kappa0 = 0.4, kappa1 = 0.6,
                       props = c(0.6, 0.3, 0.1), raters = 3)

---

Interpret Kappa Value

Description

Interpret the strength of agreement based on Landis and Koch criteria.

Usage

interpret_kappa(kappa)

Arguments

kappa

Numeric value of Kappa statistic.

Value

A named list with level, description, and color code.

Examples

interpret_kappa(0.3)
interpret_kappa(0.75)

---

Fixed N Analysis for Kappa Statistic

Description

Given a fixed sample size, estimate the lower confidence bound. Wraps kappaSize::FixedNBinary / FixedN3Cats / FixedN4Cats / FixedN5Cats.

Usage

kappa_fixed_n(n, kappa0 = 0.4, props = c(0.5, 0.5), alpha = 0.05, raters = 2)

Arguments

n	Sample size.
kappa0	Anticipated value of kappa.
props	Category proportions.
alpha	Significance level.
raters	Number of raters.

Value

List with kappaSize raw result and parameters.

---

Print Method for cate_analysis Objects

Description

Print Method for cate_analysis Objects

Usage

## S3 method for class 'cate_analysis'
print(x, ...)

Arguments

x	An object of class cate_analysis.
...	Additional arguments.

---

Print Method for cate_design Objects

Description

Print Method for cate_design Objects

Usage

## S3 method for class 'cate_design'
print(x, ...)

Arguments

x	An object of class cate_design.
...	Additional arguments.

---

Print Method for cate_fixed_n Objects

Description

Print Method for cate_fixed_n Objects

Usage

## S3 method for class 'cate_fixed_n'
print(x, ...)

Arguments

x	An object of class cate_fixed_n.
...	Additional arguments.

---

Run CATE Shiny Application

Description

Launch the Shiny application for design and analysis of consistency tests based on Kappa statistic with categorical responses.

Usage

run_cate_app(
  port = getOption("shiny.port"),
  launch.browser = getOption("shiny.launch.browser", interactive()),
  host = getOption("shiny.host", "127.0.0.1")
)

Arguments

port	The TCP port for the application. Defaults to random available port.
launch.browser	Logical. Whether to launch browser automatically.
host	The IPv4 address to listen on.

Value

A Shiny application object (invisible).

Examples

if(interactive()){
run_cate_app()
}

---

Summary Method for cate_analysis Objects

Description

Summary Method for cate_analysis Objects

Usage

## S3 method for class 'cate_analysis'
summary(object, ...)

Arguments

object	An object of class cate_analysis.
...	Additional arguments.

---