Comparing means at each timepoint wit... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Peggy Clements posted on Wednesday, September 24, 2014 - 4:32 pm
I just received the following reviewer comment on a paper that uses LGM to test the effectiveness of an intervention:

"...it is possible to compare the (least-square) means at each time point within the context of the overall LGM, and it is commonly done at each time point, much like follow-up tests of a significant interaction."

Treatment is a significant predictor of the slope. Eyeballing the means at each time point suggest that the difference between the treatment and control groups may not have emerged until the end of the study.

(Complicating matters -- this was an underpowered cluster randomized study; I used type=complex in Mplus V5 to adjust the standard errors given the nested data structure. In the manuscript, I present a regression analyses that I conducted using mplus and type=complex to compare the means at one of the time points, but the reviewer suggests that I need to do it in the context of the LGM so that I can compare the means at each of the timepoints.)

Would you please point me to a reference, page in the users' manual, or a post on the discussion board that would show me how to do this.

Thanks very much!
 Bengt O. Muthen posted on Wednesday, September 24, 2014 - 4:38 pm
I am not sure, but it sounds like the reviewer wants to see a comparison of the sample means and the LGM-estimated means at each time point. This is obtained by requesting RESIDUAL in the OUTPUT command. If you want it for the control and treatment group separately, you need to run the model as a 2-group analysis.
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: