Message/Author |
|
|
What are effect sizes for mediation effects and indirect effects in MPlus? I have tested a hypotheses about mediation in MPlus by path analysis. A model was computed that includes two mediated effects and one indirect effect (no direct effect of x on y). A reviewer asks me to give effect sizes. What are meaningful effect sizes of the mediation effects and the indirect effect? |
|
|
Usually, relating the indirect effect to the SDs of the exposure and outcome variables suffices. This is also what you get from standardized indirect effects suffice. These are given by Mplus. |
|
|
Following up the Juul's question, how does one calculate effect sizes for moderated mediation effects in Mplus? I have 5 IVs, 5Dvs, and 1 Mediator. Gender (dichotomous) is the moderator. The MODEL CONSTRAINT output indicates non-significance. However, how do I compute the effect sizes? |
|
|
This general analysis question is suitable for SEMNET. |
|
Back to top |