BR4varMCA: Computes pseudo Bootstrap ratios from pseudo-F for variables...

In HerveAbdi/data4PCCAR: Some data sets and R-functions for PCA and CA to accompany Abdi & Beaton (to appear, 2023) Principal Component and Correspondence Analysis with R

Computes pseudo Bootstrap ratios from pseudo-F for variables in MCA.

Description

BR4varMCA Computes pseudo Bootstrap ratios from pseudo-F for variables in MCA.

Usage

BR4varMCA(BrLevels, wJ, nIter = 1000)

Arguments

BrLevels	The bootstrap ratios for the variables (i.e., the output from InPosition::epMCA.inference.battery).
wJ	the masses (i.e., the center of gravity of the rows) for the columns, typically obtained from the output of ExPosition as wJ = 1 / resMCA$ExPosition.Data$W.
nIter	(Default: 1000) the number of bootstrapped iterations used to compute the original Bootstrap ratios.

Details

The idea here is to get a statistics that is commensurable across designs. So, the BRs from the levels are re-combined to give a pseudo-F that tests if the levels of the variables are reliably different for a given factor. The probability associated to the pseudo-F is then used to compute a pseudo-BR whose value will then have the same probability as the pseudo-F.

Value

A list with 6 elements

• "pseudoBR.pos" The positive pseudo BR ratios (i.e., BRs indicating differences between levels)

• "pseudoBR" BR ratios matching the probability of their F, could be positive (indicating differences between levels), or could be negative (indicating similarities between levels)

• F4VarThe F from the ANOVA testing the differences between the levels of the qualitative variable.

• df4VarThe degrees of freedom for the F from the ANOVA for the differences between the levels of the qualitative variable.

• pF4VarProbability associated the F's.

Author(s)

Hervé Abdi

Examples

## Not run: 
if(interactive()){
 #EXAMPLE1
 }

## End(Not run)

HerveAbdi/data4PCCAR documentation built on Sept. 11, 2022, 4:19 p.m.