Computes approximate relative weights for correlated predictors in a multiple
regression with one criterion.
Usage
relative_weights(predictor_cor, criterion_cor)
Arguments
- predictor_cor
Predictor correlation matrix.
- criterion_cor
Vector of predictor-criterion correlations.
Value
A data frame with raw and rescaled relative weights.
References
Johnson, J. W. (2000). A heuristic method for estimating the relative weight
of predictor variables in multiple regression. Multivariate Behavioral
Research, 35, 1-19.
Examples
# Literature: Johnson (2000).
relative_weights(matrix(c(1, .30, .30, 1), 2, 2), c(.40, .30))
#> predictor raw_weight rescaled_weight percent_of_r2
#> 1 1 0.1328022 0.1328022 67.89326
#> 2 2 0.0628022 0.0628022 32.10674
