# Tag Info

0

The total indirect effect is actually the sum of all specific indirect effects, or, more simply the total effect between $X$ and $Y$ less the direct effect between $X$ and $Y$. You can obtain barely significant specific indirect effects (which I guess happened) and when you add them up, it can reach the significance threshold.

0

What if one can develop on function of IV using $f(x_1,x_2,x_3,x_4) =b_1x_1+b_2x_2+b_3x_2+b_4x_4$ obtained from previous multiple regression model from SPSS. This can be computer using data transformation. I believe the function obtain will be the weighted result of X which is one IV then you can use Hayes processor to obtain moderate mediation of 1IV,1 ...

0

I would go for multigroup analysis. It is not clear how or where in your model the interaction should take place.

1

If you have no theoretical model to support your analysis, reporting the saturated model without fit statistics seems the best option. Removing non-significant results will gives you artificially good fit statistics, just because you developed the model in a statistics fashion (removing non significant path) rather than a theoretical one (justified ...

0

I think I just figured it out. I have inconsistent mediation - Equivalence of the mediation, confounding and suppression effect MacKinnon et al., 2000 found here: http://europepmc.org/article/PMC/2819361

0

I'm going to assume that you are running a series of simple mediation models -- one for each of your predictor variable (X1-X6). If so, then you could present the result of your mediation analyses in a single table and write about them in the results section one predictor at a time. For example: "The indirect effect of X1 via Mediator is statistically ...

1

If anyone looks at this post in the future, I have collected answers from experts, so I'm all set. The proportion mediated is negative if the direct and indirect effects have opposite directions. It makes more sense to discuss the proportion mediated if the direct is the same for both. If the outcome is binary, the ACME and ADE are absolute percent changes, ...

1

Structural equation modeling and linear regression are related, but different, analyses. If I had to guess here, lavaan is allowing variables to be measured with error by default even though you're specify regression-type models. An assumption of linear regression models, like those used for the PROCESS macro, is that all variables are measured without error....

0

I am conducting mediation analysis with 3 variables. X,Y,and M. There are many different possible outcomes and I am having difficulty drawing conclusions. I understand how to determine whether a variables is mediating etc..i just don’t know what exactly Hayes process model 4 test of X by M interaction is saying. The relationships between all variables is ...

0

Unfortunately, a group-level variable cannot be an outcome in lme4 models. Multilevel models require that the outcome be measured at the lowest level of the data hierarchy. The multilevel model partitions the variance in the outcome across the various levels - within-group, between-group, etc. Predictors at each level can be used to explain variance at their ...

Top 50 recent answers are included