Informative Nonparametric Bootstrap Test with Pooled Resampling.

npboottprmFBar

The goal of ‘npboottprmFBar’ is to implement the nonparametric bootstrap test with pooled resampling method (as presented in Dwivedi, Mallawaarachchi, and Alvarado (2017)) for informative hypothesis testing (as implemented in ‘restriktor’ and outlined in Vanbrabant and Rosseel (2020)).

Installation

You can install the released version of ‘npboottprmFBar’ from CRAN:

install.packages("npboottprmFBar")

To install the development version of ‘npboottprmFBar’ from GitHub, use the devtools package:

# install.packages("devtools")
devtools::install_github("mightymetrika/npboottprmFBar")

An iris example

The following example demonstrates how to use the bootFbar() function to conduct an informative hypothesis test.

library(npboottprmFBar)

res <- bootFbar(data = iris, formula = Sepal.Length ~ -1 + Species,
                grp = "Species",
                constraints = 'Speciessetosa < Speciesversicolor < Speciesvirginica',
                nboot = 10, conf.level = 0.95, seed = NULL, na_rm = FALSE)

paste0("Type B Test: ", res$pvalueB)
#> [1] "Type B Test: 1"
paste0("Type A Test: ", res$pvalueA)
#> [1] "Type A Test: 0"

The non-significant Type B test followed by the significant Type A test is evidence in favor the order-constrained hypothesis

References

Dwivedi AK, Mallawaarachchi I, Alvarado LA (2017). “Analysis of small sample size studies using nonparametric bootstrap test with pooled resampling method.” Statistics in Medicine, 36 (14), 2187-2205.

Vanbrabant, L., & Rosseel, Y. (2020). An Introduction to Restriktor: Evaluating informative hypotheses for linear models. In R. van de Schoot & M. Miocevic (Eds.), Small Sample Size Solutions: A Guide for Applied Researchers and Practitioners (1st ed., pp. 157 -172). Routledge. https://doi.org/10.4324/9780429273872-14