Package {PropTestR}

Compare two independent proportions

Description

Provides a unified interface for common tests and confidence intervals for two independent binomial proportions.

Usage

compare_2prop(
  x1,
  n1,
  x2,
  n2,
  method = "all",
  alternative = c("two.sided", "less", "greater"),
  margin = 0,
  conf.level = 0.95
)

Arguments

Value

An object of class 'prop_compare' containing standardized results.

Examples

result <- compare_2prop(
  x1 = 82,
  n1 = 100,
  x2 = 75,
  n2 = 100,
  method = c("wald", "score", "fisher")
)
print(result)

Two One-Sided Equivalence Test for Two Independent Proportions

Description

Performs a confidence-interval-based equivalence assessment for two independent binomial proportions using the Miettinen-Nurminen confidence interval.

Usage

equivalence_2prop(x1, n1, x2, n2, margin, conf.level = 0.9)

Arguments

Details

Equivalence is concluded when the confidence interval for the risk difference lies entirely within the prespecified equivalence margins.

Value

A data frame containing the observed risk difference, confidence interval, equivalence margin, equivalence decision, and method.

Farrington-Manning Test

Description

Performs a Farrington-Manning score test for comparing two independent binomial proportions when the null hypothesis specifies a non-zero difference.

Usage

farrington_manning(
  x1,
  n1,
  x2,
  n2,
  margin,
  alternative = "greater",
  conf.level = 0.95
)

Arguments

Value

An object of class "htest" containing the estimated difference, confidence interval, score statistic, p-value, null difference, and alternative hypothesis.

References

Farrington CP, Manning G (1990). Test statistics and sample size formulae for comparative binomial trials with null hypothesis of non-zero risk difference or non-unity relative risk. Statistics in Medicine, 9, 1447-1454.

Examples

result <- farrington_manning(
  x1 = 82,
  n1 = 100,
  x2 = 85,
  n2 = 100,
  margin = -0.10,
  alternative = "greater"
)
print(result)

Compare paired proportions

Description

Performs McNemar's test for comparing two paired binary measurements. Both the asymptotic McNemar test and the exact binomial version are supported.

Usage

paired_2prop(b, c, n, method = c("mcnemar", "mcnemar_exact"))

Arguments

Details

The paired risk difference is calculated as (b - c) / n. For the exact method, the p-value is obtained from an exact binomial test conditional on the discordant pairs, with success probability 0.5.

If there are no discordant pairs (b + c = 0), there is no evidence of a difference between the paired proportions. In this case, the p-value is set to 1 and the test statistic is returned as NA.

Value

An object of class "htest" containing the test statistic, p-value, paired difference, discordant pairs, sample size, and contingency table.

References

McNemar Q (1947). Note on the sampling error of the difference between correlated proportions or percentages. Psychometrika, 12, 153-157.

Examples

# Continuity-corrected McNemar test
paired_2prop(
  b = 15,
  c = 5,
  n = 100,
  method = "mcnemar"
)

# Exact McNemar test
paired_2prop(
  b = 15,
  c = 5,
  n = 100,
  method = "mcnemar_exact"
)

Plot confidence intervals from a PropTestR result

Description

Produces a confidence-interval plot for the methods contained in a prop_compare object.

Usage

plot_prop_result(
  x,
  main = "Two-proportion comparison",
  xlab = "Difference in proportions"
)

Arguments

Value

Invisibly returns the supplied prop_compare object.

Examples

result <- compare_2prop(
  x1 = 80,
  n1 = 100,
  x2 = 70,
  n2 = 100,
  method = c("wald", "score", "miettinen_nurminen")
)

plot_prop_result(result)

Print a PropTestR two-proportion comparison

Description

Print a PropTestR two-proportion comparison

Usage

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

Arguments

Value

Invisibly returns 'x'.

Confidence interval for a difference of proportions

Description

Calculates a confidence interval for the difference between two independent binomial proportions using established confidence interval procedures.

Usage

proportion_ci(x1, n1, x2, n2, method = "mn", conf.level = 0.95)

Arguments

Value

A data frame containing the estimated difference, lower confidence limit, upper confidence limit, and method.

Examples

proportion_ci(
  x1 = 80,
  n1 = 100,
  x2 = 70,
  n2 = 100,
  method = "mn"
)

Effect measures for two independent proportions

Description

Calculates the risk difference, risk ratio, and odds ratio for two independent binomial proportions.

Usage

proportion_effects(x1, n1, x2, n2, conf.level = 0.95)

Arguments

Details

The risk difference is calculated as the difference between the two observed proportions. The risk ratio is calculated as the ratio of the two observed proportions, and the odds ratio is calculated from the corresponding odds.

Confidence intervals for the risk difference are calculated using the Miettinen-Nurminen method. Confidence intervals for the risk ratio are calculated using the Koopman method when the underlying procedure is available.

Value

A data frame containing the effect measure, estimate, lower confidence limit, and upper confidence limit.

Examples

proportion_effects(
  x1 = 80,
  n1 = 100,
  x2 = 70,
  n2 = 100
)

Stratified comparison of two proportions

Description

Performs a Cochran-Mantel-Haenszel test for comparing two proportions across multiple independent strata.

Usage

stratified_2prop(x1, n1, x2, n2, strata, conf.level = 0.95)

Arguments

Details

Each stratum consists of a 2 x 2 table formed from the corresponding elements of x1, n1, x2, and n2. The Cochran-Mantel-Haenszel test is then performed using mantelhaen.test.

The current implementation uses the Cochran-Mantel-Haenszel procedure for testing association across strata. The conf.level argument is retained for interface consistency, although the confidence interval returned by mantelhaen.test is determined by that function's default settings in the current implementation.

Value

A list containing the Cochran-Mantel-Haenszel test method, test statistic, p-value, confidence interval, estimated common odds ratio, and the stratified contingency tables.

Examples

x1 <- c(80, 70, 90)
n1 <- c(100, 100, 100)
x2 <- c(70, 65, 85)
n2 <- c(100, 100, 100)
strata <- c("A", "B", "C")

stratified_2prop(
  x1 = x1,
  n1 = n1,
  x2 = x2,
  n2 = n2,
  strata = strata
)

Summarize a PropTestR two-proportion comparison

Description

Summarize a PropTestR two-proportion comparison

Usage

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

Arguments

Value

The results table.

Tidy a PropTestR result

Description

Extracts the standardized results table from a prop_compare object.

Usage

tidy_prop_result(x)

Arguments

Value

A data frame containing the standardized comparison results.