8
$\begingroup$

I am trying to perform a latent class growth analysis (LCGA) and/or growth mixture models (GMMs) in R. The data I am using is an increasing number of forks of git repositories (discrete variable, not categorical), as you can see in this dataset.

I tried lavaan, which helped me fit a latent growth curve model, but not to identify latent classes. I also tried poLCA, which only works for categorical polytomous variables, hence did not suffice either.

What is the most appropriate R package for performing a latent class growth analysis on discrete variable data?

The analysis I want to do is similar to that in Qureshi & Fang (2010):

Qureshi, I., and Fang, Y. 2010. “Socialization in Open Source Software Projects: A Growth Mixture Modeling Approach,” Organizational Research Methods (14:1), pp. 208–238.

$\endgroup$
4
$\begingroup$

The OpenMx project can estimate growth mixture models, though you have to install the package from their website since it isn't on CRAN. They have examples in the user documentation (section 2.8) for how to set this up as well.

$\endgroup$
1
$\begingroup$

You have also the package Kml and Kml3d (joint trajectories) that estimate the non-parametric equivalent of a GMM. You don't get any parameters as a result of these analyses, only the classification of each observation in the classes. However, in most applications, people don't use the parameters of LCGA and GMM anyway, and it is also much more robust than those applications, in particular GMM. There are two or three publications on the packages and full R documentation.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.