|
Message/Author |
|
|
I have an 11-point pain scale, where pts who have no pain are coded 0, and those with pain are coded 1-10. This item is collected across 7 assessments. I'd like to see if there are latent growth classes for this pain variable, and it seems likely that a two part model for the pain item would be best for capturing that information given the large number of zeros at each assessment. Do you have any setups for a basic two part GMM or LCGA? Thanks! - bac |
|
|
This paper, although cross-sectional, may be helpful - see the figures for the setup: http://www.statmodel.com/download/2partfmm.pdf |
|
|
Thanks very much, Bengt - this will be very helpful. The figures illustrate the logic for the analysis and the parameter estimates, but not the Mplus Model syntax that produced them. I hope you won't mind looking at an example of my syntax if need be! Best, bac |
|
Back to top |
|
|