Marginally Interpretable Models for Clustered Observations

New paper introduces marginally interpretable transformation models for clustered observations. The marginal distributions are described by a linear transformation model and the correlations by a joint multivariate normal distribution. These models can handle skewed, bounded, and survival continuous outcomes, as well as binary and ordered categorical responses.

The models can be estimated in R using the mtram() function in the tram package.

The following code can be run to see applications of these models in R:

demo("mtram", package = "tram")
vignette("mtram", package = "tram")