shrinkage_mean: Shrinkage of the standard deviation of the mean of a set of...

In mcbeem/giftedCalcs: Psychometrics for Gifted Program Identification

Shrinkage of the standard deviation of the mean of a set of scores DEPRECATED

Description

This function calculates the shrinkage (proportional reduction in the standard deviation) of a weighted mean of assessments given the reliability coefficient of each assessment, the correlations between them, and the weights.

Usage

shrinkage_mean(rely, r, w = NA)

Arguments

rely	A vector of reliability coefficients.
r	Either a correlation matrix or a vector of unique correlations. It is recommended to specify the correlations as a matrix to avoid erronous pairings of assessment correlations with reliability coefficients and weights, since this can be confusing if the correlations are supplied as a vector. Presumes that the correlation matrix has the same order of assessments as the reliability and weight vectors.
w	A vector of weights. Will be internally normalized to sum to 1 and presumes the same order of assessments as the correlation matrix and vector of reliabilities. If omitted, it is assumed that all assessments have the same weight.

Value

The shrinkage factor; a number between zero and one representing the degree of reduction in the standard deviation.

Examples

r <- matrix(c(
  1, .4, .7,
  .4, 1, .5,
  .7, .5, 1
), 3, 3, byrow = TRUE)
shrinkage_mean(rely = c(.9, .85, .88), r = r, w = c(.5, .3, .2))

mcbeem/giftedCalcs documentation built on Nov. 29, 2024, 3:48 a.m.