eq.tost: Robust and Non-Robust Variants of the Two Independent Groups...

In cribbie/equivalencetests: Equivalence Tests

Description

This R function allows for the computation of the original Schuirmann TOST for the equivalence of two independent groups, the modified Schuirmann-Welch test (which does not require the variances to be equal), or the Schuirmann-Yuen test (if normality cannot be assumed), depending on the arguments specified. The original Schuirmann assumes equal variances and normality. The Schuirmann-Welch assumes normality but not equal variances. The Schuirmann-Yuen accounts for unequal variances and nonnormality using trimmed means and Winsorized variances.

Usage

eq.tost(x, y, ei, varequiv = FALSE, normality = FALSE, tr = 0.2,
  alpha = 0.05, na.rm = TRUE, plot = TRUE, ...)

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

Arguments

x	a numeric vector for first sample
y	a numeric vector for the second sample
ei	numeric value defining the size (half-width) of the symmetric (around 0) equivalance interval
normality	logical; If true, normality of x and y are assumed.
tr	proportion of data to trim from each tail of each distribution (i.e., symmetric trimming). When tr == 0, the standard Schuirmann test or Schuirmann-Welch is performed
alpha	the maximum allowable Type I error rate
plot	whether or not to print a graphic of the results
...	additional arguments to be passed
varequal	logical; If true, equal variances are assumed. Only applicable when tr == 0
x	object of class eq.tost

Value

returns a list

Author(s)

Rob Cribbie [e-mail]

References

Schuirmann, D. J. (1987). A comparison of the two one-sided tests procedure and the power approach for assessing the equivalence of average bioavailability. Journal of pharmacokinetics and biopharmaceutics, 15(6), 657-680. van Wieringen, K. & Cribbie, R. A. (2014). Robust normative comparison tests for evaluating clinical significance. British Journal of Mathematical and Statistical Psychology, 67, 213-230. Gruman, J., Cribbie, R. A., & Arpin-Cribbie, C. A. (2007). The effects of heteroscedasticity on tests of equivalence. Journal of Modern Applied Statistical Methods, 6, 133-140.

Examples

## Not run: 

x <- rnorm(100, 1)
y <- rnorm(100, 1)
# Original Schuirman TOST
eq.tost(x, y, ei=0.5, alpha = 0.05, varequiv=FALSE, normality=FALSE, plot=TRUE)
# Schuirmann-Welch
eq.tost(x, y, ei=0.5, alpha = 0.05, varequiv=FALSE, normality=TRUE, plot=TRUE)
# Schuirmann-Yuen TOST (default)
eq.tost(x, y, ei=0.5, alpha = 0.05, plot=TRUE)
#' # Schuirmann-Yuen TOST (10% symmetric trimming)
eq.tost(x, y, ei=0.5, varequiv=FALSE, normality=FALSE, alpha = 0.05, print=TRUE) 

## End(Not run)

cribbie/equivalencetests documentation built on May 14, 2019, 11:33 a.m.