Methodological issues for a clustered ordinal response framework

In this paper, we present alternative frameworks for clustered ordinal data concerning a specific class of models denoted as CUB. Specifically, we analyze models that contain variables measured at different levels of the hierarchy by integrating them in a multi-context or multilevel structure. After a brief review on CUB models, we describe the two approaches and introduce the methodological issues for interpreting and discussing of the results.