Fraction of explained variance attributable to the mediator

A variance share is not a causal quantity and says nothing about direction: it would report the same number if the arrow between mediator and outcome ran the other way. It is a descriptive complement to a coefficient-based proportion mediated.

Usage

Vrmed(r2_full, r2_partial)

Arguments

r2_full

R-squared including the mediator.

r2_partial

R-squared without it.

Value

List with estimate, delta_r2, r2_full, r2_partial.

Details

Formula: R2_med = (R2_full - R2_partial) / R2_full.

References

de Heus, P. (2012). Behavior Research Methods 44:213-221.

Examples

Vrmed(r2_full = 5L, r2_partial = c(1, 2, 3, 4, 5, 6, 7, 8))
#> $estimate
#> [1]  0.8  0.6  0.4  0.2  0.0 -0.2 -0.4 -0.6
#> 
#> $delta_r2
#> [1]  4  3  2  1  0 -1 -2 -3
#> 
#> $r2_full
#> [1] 5
#> 
#> $r2_partial
#> [1] 1 2 3 4 5 6 7 8
#> 
#> $method
#> [1] "Variance-based mediation share"
#> 
#> attr(,"class")
#> [1] "morie_rich_result" "list"