Growth mixture modeling in SPSS?

[Frank Furter]

I have a data set with longitudinal measurements of a biological parameter taken at similar (but not identical) time points, and I would like to identify homogeneous subsets of subjects based on their growth trajectories. It appears that this can be achieved using growth mixture modeling (GMM).

Can this be done in SPSS (I have Version 28 under Windows)? I've heard that this may be possible in Amos, but there seem to be different opinions on that. Has anyone here got experience with GMM analysis in SPSS?

[Andy W]

IMO easiest path is to pass in your data into R, see https://andrewpwheeler.com/2014/08/12/estimating-group-based-trajectory-models-using-spss-and-r/ for an example.

That is for a special library for count mixtures over time, I would recommend the flexmix library though for most problems, https://andrewpwheeler.com/2021/03/05/simulating-group-based-trajectories-in-r/.

[Frank Furter]

Thank you so much, Andy! Indeed, there appear to be several GMM packages in R, and I like the idea of passing the data to them from within SPSS if it cannot be done in SPSS without using additional software.

BTW, my data are (ratio-scale) measurements of a biological outcome measure, so I will have to use a different GMM package.