equiv_drs: Dependent Correlations

In cribbie/equivalencetests: Equivalence Tests

Description

Tests the equivalence of the correlations r12 and r13 (variable 1 is overlapping variable) in a three variable correlation matrix. Standard errors based on William's modification to Hotellings test comparing dependent overlapping correlations.

Usage

equiv_drs(dat, ei, n = NULL, alpha = 0.05)

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

Arguments

dat	an N x 3 matrix or data.frame containing raw data used to compute the correlation matrix between variables. The input may also be a 1 x 3 vector of correlations (r12, r13, and r23, respectively) and requires a sample size input (N)
ei	equivalence interval
n	sample size when dat input is a vector of correlations
alpha	desired alpha level
x	object of class equiv_drs
...	additional arguments

Value

returns a list containing the p-value, confidence interval, and statistical decision

Author(s)

Rob Cribbie [e-mail] and Alyssa Counsell [e-mail]

Examples

## Not run: 
#raw data
set.seed(1234)
dat <- cbind(rnorm(100), rnorm(100), rnorm(100)) 
equiv_drs(dat, ei = .2)

#correlations input
r12 <- cor(dat)[2,1]
r13 <- cor(dat)[3,1]
r23 <- cor(dat)[3,2]
equiv_drs(c(r12, r13, r23), ei = .2, n = nrow(dat))

## End(Not run)

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