ccor: Scale-Center (or Cohen) Correlations, rc

View source: R/ccor.R

ccorR Documentation

Scale-Center (or Cohen) Correlations, rc

Description

Estimate correlations between variables where the scale midpoint serves as the reflection point for estimating correlations.

These are equivalent to Cohen's (1969) rc correlations, which are invariant across potentially arbitrary item- or variable-reversals.

Usage

ccor(x, mid = 3) #e.g., midpoint of standard 1-5 Likert scale is 3

Arguments

x

matrix to be used for correlations

mid

Scale midpoint (default if not given = 0)

margin

Specify whether correlations are of rows=1 or columns = 2 (default is columns)

Details

Function computes Cohen-adjusted correlations (or avg. product of standardized deviations-from-scale-center)

Value

Scale-center correlations (r_c)

Note that these are NOT the same as regular correlations done on 'scale-centered' scores. See Wood (202x) for details on how to interpret. Additionally: this script is not particularly friendly toward missing values.


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