Moderation PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Luci M. posted on Monday, May 09, 2011 - 10:06 pm
Hi!

I have a follow up question. I ran my model in MPlus and got a couple of significant interaction effects. I would like to now find the value of the slope when age is at certain values (e.g., minimum and maximum age in my sample). How can I do this in MPlus?

The way I've been doing this so far is to add or substract manually the coefficient from the interaction term to the estimate of the conditional variable (the estimate of my predictor variable when age is at the mean).

I would also like to plot the moderation effect and if possible the slopes at different ages. I have previously found a program that does this for age at +/-1 SD, but I would prefer doing this in MPlus with the slopes for different ages instead.

Thank you very much!
 Linda K. Muthen posted on Tuesday, May 10, 2011 - 6:05 am
See Slides 170-171 of the Topic 3 course handout for a suggestion of how to interpret a latent variable interaction.
 Luci M. posted on Tuesday, May 10, 2011 - 5:01 pm
I looked at the slides and while they are helpful in presenting one way to interpret interaction effects, I still struggle with finding a way to interpret the effect of my predictor on my dependent variable when age is at different values. This way, I can actually show at what age the relation starts to become significant.

I have centered age, and then created additional variable with age at certain values by substracting the specific age from the raw variable 'age'. However, I don't know if I need to compute additional product terms with age at the specific value (let's say 7, 10, 12, 14, 16). If so, I am not sure if I need to do any transformations with my predictor variable (which I have mean centered for now).

Ideally, I would like to use the unstandardized regression coeff. as to me they are more intuitive. Thank you very much for all your help.
 Bengt O. Muthen posted on Tuesday, May 10, 2011 - 5:29 pm
Perhaps you want to look at the book by Aiken & West on how to interpret and display interaction effects. Or, see if SEMNET readers have suggestions.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: