Fits \(Y = i_1 + cX\), \(M = i_2 + aX\) and
\(Y = i_3 + c'X + bM\), and reports each of the four conditions
separately rather than collapsing to a verdict. Requiring a
significant total effect (step 1) is now regarded as too strong:
opposite-signed direct and indirect paths cancel, and that rule
discards exactly those cases.
Usage
morie_baron_kenny(y, x, m, alpha = 0.05)
Arguments
- y, x, m
Outcome, treatment and mediator vectors.
- alpha
Significance level for each step.
Value
Named list with c, a, b,
c_prime, indirect, proportion_mediated,
se, p, steps, mediation, n.
Details
Mirrors morie.fn.bkmed.
References
Baron RM & Kenny DA (1986). The moderator-mediator
variable distinction in social psychological research: conceptual,
strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51(6), 1173-1182.
Examples
set.seed(1)
x <- rnorm(200)
m <- 0.8 * x + rnorm(200)
morie_baron_kenny(0.3 * x + 0.6 * m + rnorm(200), x, m)$mediation
#> [1] "partial"
