scalebyN: Scale by N

In funfield-lab/fancyr: Fancy Statistics for Correlational (r) Analyses

Scale by N

Description

Estimate z-scores for column variables using N rather than standard N-1

Usage

scalebyN(x, center = T, scale = T)

Arguments

x	matrix containing set of scores to be transformed
center	mean-center scores (run ?scale for more information
scale	divide by standard deviation (run ?scale for more information

Details

Why do this? Because it is easier to show that correlations are equivalent to (1) the average product of zX*zY scores (Cohen, Cohen, Aiken, & West, 2003; Eq. 2.3.1) or (2) the average squared difference of zX and zY scores (Cohen, Cohen, Aiken, & West, 2003; Eq. 2.2.4) when using z-scores computed by using N rather than N-1. Indeed, when done so, the averages of these product decompositions become strictly equivalent.

Value

Standardized scores, calculated using N rather than N-1 for standard deviations

funfield-lab/fancyr documentation built on Nov. 21, 2023, 2:42 p.m.