
Message/Author 

jan mod posted on Wednesday, March 21, 2018  5:13 am



I am using the TNIE in mediation analysis. I have a continuous X, binary M and continuous Y. I give two values of the continuous X (1 SD and +1 SD) that is hypothesized to be mediated by binary mediators. Is the following correct? 1) TNIE has two parts in the equation: the first part is the mean of Y when X=1 (in this case +1 SD) while M varies as it would be for X=1. The second term in the equation is the mean of Y when X=1 while M varies as it would be if X=0 (in this case 1 SD). 2) The effect of TNIE shows me then how much Y increases or decreases when X goes from 1 SD to + 1SD as an indirect effect. 


Looks right. Our book Regression and Mediation Analysis using Mplus has an example of this case (although with a binary X) in Section 4.3.1. 

Back to top 

