Interaction between manifest and late... PreviousNext
Mplus Discussion > Structural Equation Modeling >
 michele roccato posted on Thursday, December 11, 2008 - 7:58 am

I have to model an interaction between a manifest and a latent variable in a Sem. In the first version of the paper I modeled it as the product between the manifest variable and the factorial score of the latent variable. However, an anonymous reviewer asked me to model it as a latent variable.

I would prefer not to follow this suggestion. Could you please suggest me some papers I could cite to support my decision not to follow this suggestion?

I thank you in advance

 Linda K. Muthen posted on Thursday, December 11, 2008 - 8:15 am
I would have to agree with the reviewer. I don't know of any papers that support your approach. You can use the XWITH option to model the interaction.
 michele roccato posted on Monday, December 15, 2008 - 6:56 am
 lotti posted on Thursday, April 04, 2013 - 8:16 am
Dear Prof. Muthén,

I'm dealing with the fllowing latent interaction model:

Rob2 = intercept + B1(rob1)+B2(age)+B3(age2)+B4(ageXrob1)+B5(rob1Xage2)+e

Rob1 and Rob 2 are both latent variables measured by five indicators at two time points, representing performance. age is subjects' chronological age, and age 2 is age squared.
Model is identified following standard procedures, and both interaction terms are defined as random slopes (option "random" enabled).

I would like to obtain standardized parameters estimates.
Can you, please, to suggest a way to expand the formulation you offered in your webnote "latent Variable interaction", Muthén, 20 September, 2012, paragraph 1.4.
I thank you in advance.
 Bengt O. Muthen posted on Thursday, April 04, 2013 - 1:18 pm
Why do you need to standardize?
 lotti posted on Thursday, April 04, 2013 - 1:57 pm
Dear Prof. Muthén,

first of all thank you for your quick answer.

Since B1 represents the stability of performance, I would like to have the standardized value of B1 at diferent levels of moderators (age and age2). In this regard, following your webnote, I was able to derive the moderating function for unstandardized terms and to build a graph. However, the use of standardized values could make the interpretation of changes in the stability of performance more readily understandable. What do you think?
Thank you so much for your availability and your time.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message