conf.reg {MANOVA.RM}

The conf.reg() function calculates confidence regions for contrasts in multivariate factorial designs. In the two-dimensional case, confidence ellipsoids can be plotted via the generic plot() function.

Description

The conf.reg() function calculates confidence regions for contrasts in multivariate factorial designs. In the two-dimensional case, confidence ellipsoids can be plotted via the generic plot() function.

Usage

conf.reg(object, nullhypo)

Arguments

object

A MANOVA object.

nullhypo

In scenarios with more than one factor, the null hypothesis, i.e., the contrast of interest must be specified.

Value

A confreg object containing the following components:

center

The center of the confidence ellipsoid.

scale

The scaling factors for the axis of the confidence ellipsoid calculated as \sqrt{\lambda*c/N}, where \lambda are the eigenvalues, c denotes the bootstrap quantile and N is the total sample size. See Friedrich and Pauly (2018) for details.

eigenvectors

The corresponding eigenvectors, which determine the axes of the ellipsoid.

References

Friedrich, S., and Pauly, M. (2018). MATS: Inference for potentially singular and heteroscedastic MANOVA. Journal of Multivariate Analysis, 165, 166-179.

Examples

data(EEG)
EEG_mod <- MANOVA(resp ~ sex * diagnosis, 
                    data = EEG, subject = "id", resampling = "paramBS", 
                    alpha = 0.05, iter = 10, CPU = 1)
conf.reg(EEG_mod, nullhypo = "sex")