Homology test / emergent effects PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 Hadar Nesher Shoshan  posted on Wednesday, September 12, 2018 - 3:58 am
Dear Mplus team,
I try to test for homology or emergent effects in a multilevel SEM. following Hülsheger et al., 2014 (p. 1119), I tried to do it with Mplus.

However, I cannot find the procedure in the manual, or understand how was the significance test of the between-within paths done in Mplus.

I have tried to constrain the between and within paths to be the same, and compare the models with the Satora Bentler chi square test. Could this be appropriate?
Thank you very much
 Bengt O. Muthen posted on Wednesday, September 12, 2018 - 4:53 pm
I am not familiar with this article - please send full reference.
 Hadar Nesher Shoshan  posted on Wednesday, September 12, 2018 - 11:33 pm
Thank you. The full reference is: Hülsheger, U. R., Lang, J. W., Depenbrock, F., Fehrmann, C., Zijlstra, F. R., & Alberts, H. J. (2014). The power of presence: The role of mindfulness at work for daily levels and change trajectories of psychological detachment and sleep quality. Journal of Applied Psychology, 99(6), 1113-1128.
The relevant paragraph is on page 1119: "Yet, as outlined in the introduction, we were not only interested in independent effects at both levels of analysis but also in whether effects differ significantly between levels of analysis. The question was thus whether an emergent effect exists (Chen et al., 2005; Enders & Tofighi, 2007; Lüdtke et al., 2008), for instance, if the mindfulness–psychological detachment relationship is stronger at the between-person than at the within-person level of analysis. To test for emergent effects, we used the approach described by Muthen and Muthen (1998 –2012, pp. 262–263; also see Ludtke et al., 2008). In doing so, we estimated additional values indicating the differences between the Level 1 and Level 2 effects (aw vs. ab, bw vs. bb, and cw vs. cb; see Figure 1)."
 Bengt O. Muthen posted on Friday, September 14, 2018 - 1:36 pm
You mention pages 262-263 but not sure what you refer to - can you please give me the corresponding pages for the Version 8 User's Guide that is online?
 Hadar Nesher Shoshan  posted on Friday, September 14, 2018 - 2:01 pm
Thank you. These are the pages cited of Hülsheger et al., 2014. The corresponding pages in version 8 are 273-275
 Bengt O. Muthen posted on Sunday, September 16, 2018 - 1:52 pm
Yes, you can use Satorra- Bentler difference testing but perhaps a simpler approach is to label the a, b, c parameters in the Model command and then use Model Constraint to test their differences. E.g.

Model Constraint:

New(adiff bdiff cdiff);

adiff = ab - aw;
bdiff = bb - bw;
cdiff = cb - cw;

This gives you the significance of adiff, bdiff, and cdiff.
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