Should I exclude some cases? PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Boram Lee posted on Wednesday, January 04, 2017 - 5:03 pm
Hello. I'm a doctoral student in the field of child development and family studies.

I have a question about LCGA solution.
I'm in the process writing a doctoral dissertation using Growth Mixture Modiling using LCGA.

The main purpose of my study is to examine maternal puerperal depression trajectories and classify groups according to their depression change tendency.

There are three time point information about maternal depression(at pregnancy, 1-year after the birth, and 2-year after the birth).

I could classified two groups according to the type of maternal depression tajctories, 'depression deepening group'(10%: 70 cases) and 'depression relieving group'(90%, 647cases), using LCGA. Two group solution showed the best fitness of AIC, BIC, Entropy, and ALMR.

The problem is that my commitee members suggested that I should exclude 8 cases that could be classified in 3 class solution by LCGA. Their (8 cases: 0.1% of samples) depression trajectory was low level maintenance tendency.

What do you think? Should I exclude 8 cases when I identify the differences between the two groups- depression deepening group and depressiojn relieving group?

I would appreciate your suggestions.
Thank you.
 Linda K. Muthen posted on Thursday, January 05, 2017 - 6:11 am
This type of general question is more appropriate for a general discussion forum like SEMNET. We try to keep questions specific to Mplus on this forum.
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: