often the justification for removing (trimming) or recoding (winsorizing) observations whose values lie outside a specified … winsorizing and trimming are only solutions for a narrow class of measurement error processes. Indeed, for the measurement error … results from the Census PUMS data and 2001 CPS data demonstrate the fragility of trimming and winsorizing as solutions to …

Two approaches for model-based clustering of categorical time series based on time- homogeneous first-order Markov … chains are discussed. For Markov chain clustering the in- dividual transition probabilities are fixed to a group …-specific transition matrix. In a new approach called Dirichlet multinomial clustering the rows of the individual transition matri- ces …

the hierarchy on the variance decomposition and the clustering effects have not been well documented. This paper discusses … how omitting one level in three-level models affects the variance decomposition and clustering in the resulting two …