0

I have the following mediation analysis with X as the independent variable, Y as the dependent variable and three mediators (M1, M2, M3).

An analysis of the direct relationships shows the following results:

X -> M1 = p-value: 0.016 & beta-coefficient: -0.197;

X -> M2 = p-value: 0.001 & beta-coefficient: 0.274;

X -> M3 = p-value: 0.055 & beta-coefficient: 0.185;

M1 -> Y = p-value: 0.019 & beta-coefficient: -0.277;

M2 -> Y = p-value: 0.199 & beta-coefficient: 0.104;

M3 -> Y = p-value: 0.009 & beta-coefficient: 0.333;

The mediating effects, however, are all not significant:

X -> M1 -> Y = p-value: 0.068 & beta-coefficient: 0.055;

X -> M2 -> Y = p-value: 0.202 & beta-coefficient: 0.028;

X -> M3 -> Y = p-value: 0.080 & beta-coefficient: 0.062;

The confusing part, however, is that when I calculate the total indirect effect the relationships seems to be significant =

p-value: 0.003 & beta-coefficient: 0.145

How should I interpret these findings?

Elisa
  • 21
  • 1
  • See https://stats.stackexchange.com/questions/3549/why-is-it-possible-to-get-significant-f-statistic-p-001-but-non-significant-r. – whuber Dec 02 '20 at 16:08
  • See also https://stats.stackexchange.com/questions/502643/mediation-how-can-the-total-indirect-effect-be-significant-when-the-indirect-ef/506135#506135 The answer is exactly what happened here. – POC Jan 23 '21 at 00:39

0 Answers0