bootTest

0th

Percentile

Bootstrap test for discretized normality

bootTest is a bootstrap test for whether an ordinal dataset is consistent with being a discretization of a multivariate normal dataset.

Usage
bootTest(my.data, B = 1000, verbose = TRUE)
Arguments
my.data

A dataset containing ordinal data. Must contain only integer values.

B

Number of bootstrap samples.

verbose

If true, bootstrap progress is printed to the console.

Value

p-value associated with the underlying normality hypothesis.

References

Nj<U+00E5>l Foldnes & Steffen Gr<U+00F8>nneberg (2019) Pernicious Polychorics: The Impact and Detection of Underlying Non-normality, Structural Equation Modeling: A Multidisciplinary Journal, DOI: 10.1080/10705511.2019.1673168

Aliases
  • bootTest
Examples
# NOT RUN {
set.seed(1)
norm.data <- MASS::mvrnorm(300, m=rep(0,3), 
Sigma=cov(MASS::mvrnorm(15, mu=rep(0,3), Sigma=diag(3))))
disc.data <- apply(norm.data,2,  cut, 
breaks = c(-Inf, 0,1, Inf), labels=FALSE)# normal data discretized
pvalue <- bootTest(disc.data, B=500)
#no support for underlying non-normality
# }
Documentation reproduced from package discnorm, version 0.1.0, License: GPL (>= 2)

Community examples

Looks like there are no examples yet.